Skills, not just Diplomas

DIREC TIONS IN DE VELOPMENT
Human Development
Skills, Not Just Diplomas

Managing Education for Results
in Eastern Europe and Central Asia
Lars Sondergaard and Mamta Murthi

with Dina Abu-Ghaida, Christian Bodewig, and Jan Rutkowski
Managing Education for Results
in Eastern Europe and Central Asia
Lars Sondergaard and Mamta Murthi
with Dina Abu-Ghaida, Christian Bodewig, and Jan Rutkowski
© 2012 The International Bank for Reconstruction and Development / The World Bank
1818 H Street NW
Washington DC 20433
Telephone: 202-473-1000
Internet: www.worldbank.org
All rights reserved
1 2 3 4 :: 14 13 12 11
This volume is a product of the staff of the International Bank for Reconstruction and
Development / The World Bank. The findings, interpretations, and conclusions expressed
in this volume do not necessarily reflect the views of the Executive Directors of The World
Bank or the governments they represent.
The World Bank does not guarantee the accuracy of the data included in this work. The
boundaries, colors, denominations, and other information shown on any map in this work
do not imply any judgement on the part of The World Bank concerning the legal status of
any territory or the endorsement or acceptance of such boundaries.
Rights and Permissions

The material in this publication is copyrighted. Copying and/or transmitting portions or all
of this work without permission may be a violation of applicable law. The International
Bank for Reconstruction and Development / The World Bank encourages dissemination of
its work and will normally grant permission to reproduce portions of the work promptly.
For permission to photocopy or reprint any part of this work, please send a request with
complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive,
Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www
.copyright.com.
All other queries on rights and licenses, including subsidiary rights, should be addressed to
the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433,
USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org.
ISBN: 978-0-8213-8096-3
eISBN: 978-0-8213-8097-0
DOI: 10.1596/978-0-8213-8096-3
Library of Congress Cataloging-in-Publication data has been requested.
Cover photo: The graduating class of 2007, Alexandru Ioan Cuza University, Iaşi, Romania,
march to their commencement ceremony. © Mediafax Foto/Liviu Chirica.
Contents
Foreword xv
Acknowledgments xix
Abbreviations xxi
Overview 1
The Skills Challenge 2
Why Are Skills an Emerging Problem if
Education Systems Are Delivering? 4
Priority Areas for Action 7
Managing Education Systems for Results 9
Build the Foundations of Adult Learning Systems 14
Conclusion 14
Notes 15
Chapter 1 The Demand for Skills in ECA 17

Background: The Demand for Highly Skilled
Labor in the Global Knowledge Economy 18
Demand for Skilled Labor Has Risen in the
ECA Region 20
Unemployment Patterns in ECA Countries
Confirm the Demand for Skilled Labor 25
v
vi Contents
Wages Have Risen for Skilled Labor 31

Lack of Needed Skills Is Impeding
Enterprise Growth 36
Skills Mismatch in the ECA Region 39
Summary 41
Notes 43
Chapter 2 Education and the Supply of Skills to the

ECA Market 47
Background: The Global Knowledge Economy
Requires Lifelong Learning 48
Formal Education in ECA Countries: High
Attainment and Good Quality Relative
to Current Income Levels 50
Why Are Skills Emerging as a Problem if
Students May Not Be Acquiring the Right Skills 63
Adult Learning Is Limited in the Region 65
Summary 77
Annex 2A: Education Systems in ECA Today 80
Notes 87
Chapter 3 Resolving the Skills Shortage in the

ECA Region: A Policy Framework 89
Operating in the Dark: Ministries Know
Too Little to Effectively Manage the
Education Sector 90
Legacy of Central Planning 96
Inefficient Use of Funds 103
Addressing the Skills Challenge 108
Summary 111
Notes 112
Chapter 4 Managing for Results at the Pre-University

Level of Education 115
Track Student Learning and Employment
Outcomes 116
Expand Autonomy in Exchange for
Accountability for Results 120
Contents vii
Improve the Efficiency of Resource Use 126

Summary 136
Notes 138
Chapter 5 Managing for Results in the Tertiary

Education Sector 139
Introduce Learning Assessments and Track
Employment Outcomes 140
Strengthen Accountability 152
Introduce Performance-Based Financing and
Encourage Private Funding Resources 159
Summary 162
Notes 163
Chapter 6 Advancing Adult Learning in ECA 165

Building the Foundations for Adult
Learning Systems 166
Promote Autonomy and Accountability
of both Public and Private Providers 172
Ensure the Efficiency of Sector Financing 175
Continued Government Role in Retraining
and Education for the Unemployed 183
Priorities for Adult Education and Training
Systems in ECA Countries 187
Summary 191
Notes 191
Chapter 7 Extended Summary: The Path for Education

Reforms in the ECA Region 193
The Skills Challenge in the ECA Region 194
Why Are Skills an Emerging Problem if
Priority Areas for Action 203
Managing Education Systems for Results 207
Build the Foundations of Adult Learning Systems 219
Summary 220
Notes 222
References 225
viii Contents
Boxes
2.1 Defining “Skills,” “Competencies,” and Other Terms 49
2.2 Testing for Reading Competency in the Programme
for International Student Assessment (PISA) 55
2.3 Turkey: Many Students Fail to Acquire Even the
Most Basic Proficiencies 57
2.4 Russia: Building Higher-Order Skills Is Proving Difficult 59
2.5 The History of Vocational Education in the ECA Region 64
2.6 Two Forms of Adult Learning 69
3.1 The Value of Standardized External Student
Assessments in ECA Countries 94
3.2 Detailed Regulations Set Norms Even for
Education Facilities in the ECA Region 99
4.1 Using Data to Measure Gaps and Design Better
Policies: Three Examples 118
4.2 Eliminating Class Size Norms in Denmark 124
4.3 Rethinking Secondary Education 125
4.4 Per Student Funding Formula: Recognizing the
Varied Costs of Providing Education 128
4.5 Rationalizing Class Sizes in Bulgaria 133
4.6 Developing New Teacher Policies 135
5.1 The Difficulty of Measuring Competencies at
the Tertiary Level 142
5.2 Decentralization of the University Sector in Romania 146
5.3 Introducing Businesslike Leadership and
Management in Lithuanian Universities 149
5.4 Using External Watchdogs to Shed Light
on Integrity Problems 155
5.5 Two Approaches to Strengthening Accountability 157
6.1 Lifelong Learning Strategy of the Czech Republic 167
6.2 Policy Tools for Advancing Adult Learning in the
United Kingdom 169
6.3 Market Failures that Impede Adult Learning 176
6.4 Supporting Small and Medium Enterprises to
Participate in Training 181
6.5 Second-Chance and Remedial Education 184
7.1 Higher-Order Skills for the World of Work in
the 21st Century 199
7.2 A Large Proportion of Students Are Failing 201
Contents ix
7.3 Options for Making Schools More Accountable for

Learning Outcomes 213
Figures
O.1 Analysis of Reading Competency of 15-Year-Old
Students on the PISA 2009 3
O.2 Distribution of Firms in ECA Region that Consider
Worker Skills a “Major” or “Very Severe”
Constraint, 2008 4
O.3 Primary School Student-Teacher Ratios in ECA
Compared to Other Regions of the World, 1990–2008 8
O.4 Status of Measuring and Using Data on Student
Learning Outcomes in the ECA Region, 2009 10
1.1 Trends in Routine and Nonroutine Tasks of U.S.
Labor Force, 1960–2002 19
1.2 Worker Responses to Survey on Working Conditions
in Europe, 2005 20
1.3 Job Creation and Destruction Rates for Selected
Occupations in Georgia, 2007 22
1.4 Job Creation and Destruction Rates for Selected
Occupations in Ukraine, 2007 23
1.5 Changes in Occupational Share of Total Employment
in Selected ECA Countries, Various Years 24
1.6 Yearly Outflow Rates from Employment to Jobs
by Occupation in Two Subregions of ECA, 2006 27
1.7 Unemployment-to-Vacancy Ratio by Occupation in
Croatia, the Czech Republic, and Ukraine, Various Years 28
1.8 Labor Shortages and Surpluses by Level of
Educational Attainment, Selected ECA Countries 29
1.9 Returns to Schooling in Selected ECA Transition
Countries, Various Years 32
1.10 Wage Premia by Level of Education in
Hungary, 1986–2004 33
1.11 Wage Premia by Education Level in the Russian
Federation and Turkey 34
1.12 Premia by Occupation Relative to Elementary
Occupations in Various ECA Countries 35
1.13 Wage Growth by Occupation in Bulgaria
and Poland, 1996–2006 37
x Contents
1.14 Employers’ Perceptions of Worker Skills as a

Constraint to Growth, Various ECA Countries 38
1.15 Time Needed by Firms to Hire Specific Workers,
by ECA Subregion, 2005 40
1.16 Employers’ Valuation of Workers’ Knowledge
and Skills in Kazakhstan and Poland 42
2.1 Gross Enrollment Rates in Tertiary Education,
by World Bank Region 51
2.2 Assessing a Skill—Reading Performance in the
Fourth Grade: PIRLS Performance of ECA
Countries, 2001 and 2006 53
2.3 Analysis of Reading Competency of 15-Year-Old
2.4 Assessing a Competency—Reading Level at
Age 15: Share of Students Scoring Level 1 or
Below on Reading Section of PISA 2009 56
B2.3 Distribution of PISA 2006 Math Scores by
Type of School in Turkey 57
2.5 Students Aware of Bribery for Grade or Exam
in their Faculty 62
B2.5 Global Comparison of Average Vocational
Enrollment of Upper Secondary Students
by Region, 1989, 1999, and 2007 64
2.6 Share of Upper Secondary Students Enrolled in
Vocational Programs in ECA Countries over Time 66
2.7 Share of ECA Firms that Offer Formal Training
Programs for Permanent, Full-Time Employees, 2008 70
2.8 Employee- and Firm-level Data on Worker
Participation in CVET Courses, EU10 and other
EU Member States 72
2.9 Training Length and Percentage of Employees
Participating in CVET Courses, EU10 and EU27 73
2.10 ECA Firms that Offer Formal Training to Employees,
by Employee Category, 2005 74
2.11 Large and Small Firms that Offer Training to Skilled
Employees, ECA, EU, and United Kingdom, 2005 76
2.12 Adult Education and Training Program Participants
in EU, 2007 78
3.1 Educational Background of 25–34-year-olds in the
ECA Region, 2006 91
Contents xi
3.2 Learning to Use Data to Drive Education Policy 95

3.3 Status of Measuring and Using Data on Student
3.4 Relationship between Primary School Test Scores,
Adult Education Levels, and Per Student Spending
in Two Municipalities of Poland 100
3.5 Relationship between Primary School Test Scores,
Average Class Size, and Per Student Spending
in Two Municipalities of Poland 101
3.6 Dynamics of the 6- to 12-year-old Population in the
ECA Region, 1990–2006 104
3.7 Primary School Student-Teacher Ratios in ECA
Compared to Other Regions of the World,
1990–2008 105
3.8 Real Per Student Expenditure Compared to TIMSS
Math Scores in Romania, 1999–2008 106
3.9 Percentage of Students with a Teacher Over 50 Years
Old in ECA Countries, Selected Years 107
3.10 Progress Towards Results-based Education Financing
in the ECA Region, 2010 110
B4.4 Groupings of Bulgarian Municipalities, Together
with Baseline Per Student Financing Amounts
and Adjustment Coefficients, 2007 128
4.1 Comparison of Average Size of Primary School
Classes Worldwide, Various Years 131
4.2 Average Class Size in Large Primary and
Secondary Schools in Nine ECA Countries,
Various Years 132
B4.5A Estimated Number of Public School Closures in
Bulgaria from 1991–92 through 2008–09 134
B4.5B Average Class Size and Student-Teacher Ratios
in Bulgaria, 2000–08 134
5.1 Progress on Implementing Quality-Assurance
Mechanisms: The Bologna Scorecard 158
B6.3 Barriers to the Expansion of Adult Education and
Training in the ECA Region 177
6.1 Employment Placement Rates by Type of Retraining,
Turkey, 2008 186
6.2 Adult Education and Training Priorities in the
ECA Region 187
xii Contents
7.1 Assessing a Skill—Reading Performance in the

Fourth Grade: PIRLS Performance of ECA Countries,
2001 and 2006 196
7.2 Analysis of Reading Competency of 15-Year-Old
7.3 Distribution of Firms in ECA Region That
Consider Worker Skills a “Major” or “Very Severe”
Constraint, 2008 198
7.4 Worker Responses to Survey on Working Conditions
in Europe, 2005 200
7.5 Students Report Unofficial Payments Are “Usually”
or “Always” Needed in Public Technical Colleges
and Universities, 2006 203
7.6 Primary School Student-Teacher Ratios in ECA
Compared to Other Regions of the World, 1990–2008 205
7.7 Comparison of Average Size of Primary School
Classes Worldwide, Various Years 206
7.8 Percentage of Students with a Teacher Over 50 Years
Old in ECA Countries, Selected Years 207
7.9 Status of Measuring and Using Data on Student
7.10 Progress Towards Results-based Education Financing
in the ECA Region, 2009 217
Tables
1.1 Unemployment Rates by Educational Attainment and
ECA Subregion, 2006 25
2.1 Time Needed to Find First Job by Level of
Educational Attainment in Serbia and Ukraine 67
2.2 Training Participants as Share of Participants in Active
Labor Market Policies and of Total Unemployed,
EU27 and EU10, 2006 77
2A.1 Duration of Education in ECA Countries, Various Years 80
2A.2 ECA Country Results on International Assessments
since 1995 85
3.1 Information Collected from Tracer Study of Dutch
University Graduates, 2007 97
B4.1 Learning Gaps between Roma and Non-Roma
Students in Serbia, as Measured by PISA 2006
and 2009 Results 118
Contents xiii
B5.3 Summary of Changes 150

5.1 Tools for Strengthening Basic Academic and Fiscal
Integrity in University-Level Institutions 154
5.2 Three Models of Performance-Based Funding 161
6.1 Recommended Policy Framework for Adult Education
and Training in the ECA Region 189
B7.2 Proportion of 15-Year-Old Students in ECA Who
Achieved Only Basic Reading Competency
on PISA 2009 201
7.1 Information Collected from Tracer Study of Dutch
University Graduates, 2007 211
7.2 Tools for Strengthening Academic and Fiscal Integrity
in University-Level Institutions 215
Foreword
It is no secret that good education lies at the heart of economic growth

and development. At the same time, improving the quality and relevance
of education is enormously difficult not least because there is no one
single policy measure that will do so effectively. This book contributes to
our understanding of how to improve education by examining the recent
experience of the countries of Eastern Europe and the former Soviet
Union.
Fabled for uniform access and high quality of education 20 years ago,
the countries in this region have struggled to maintain their reputation.
Three factors have contributed to the slide in quality and relevance.
First, one of the legacies of central planning is that the countries pay
too much attention to the measurement of inputs into learning—such
as the number of schools and the number of teachers—and not enough
to outcomes. Indeed, they have been late in developing systems to
assess how much students are learning and whether learning is leading
to employment. In this sense, their education systems operate in the
dark, which makes policy making extremely difficult. Second, the sys-
tem of management, also a vestige of the past, limits the ability of
schools to improve the learning environment for students, as well of
municipalities that may want a different mix of programs to meet local
xv
xvi Foreword
labor market needs. As with assessment, the countries have been slow
to embrace the governance and accountability reforms that are now
part of the landscape of education systems the world over. These limita-
tions to autonomy and accountability for outcomes have reduced the
energy and the incentives for improvements from within the system,
contributing to a shortage of skills. Indeed, firm complaints about the
shortage of relevant skills for expansion and growth have risen to a cre-
scendo in most countries. Finally, the systems increasingly allocate
resources where they are not needed. For example, the sharp decline in
student numbers in the past 20 years has not resulted in a commensu-
rate decline in the number of classrooms and teachers. As a result,
resources are increasingly tied up in buildings and teachers where they
may be better spent elsewhere. Most of these limitations are found not
just in schools but extend to higher education and to training.
The result of these limitations is that the quality and relevance of
education in the countries in the region is increasingly questionable. A
large proportion of students finish lower secondary school with a mini-
mal command of literacy and numeracy. Their failure to learn even the
basics is not picked up early enough by assessment systems, nor are
schools and municipalities given sufficient incentives to ensure that all
students learn the basics. The systems are also struggling with imparting
higher order skills beyond the basics, for which well-motivated and high-
quality teachers are critically needed but which few school systems are
able to attract, given the low pay and historical overstaffing. And without
the vision or resources to make vocational and technical education an
attractive option for students—indeed, this remains one of the most
unreformed subsectors of the education system—quality has eroded and
students have shied away from this form of education, possibly contrib-
uting to the widespread shortage of vocational and technical skills now
found in these countries.
This book makes the case that improving the quality and relevance of
education requires a fundamental change of approach to education in the
countries of the region. To start with, education systems need to “turn the
lights on” and take seriously the measurement of what students actually
learn as opposed to measurement of the inputs into the education process
on the implicit assumption that learning follows. This assessment needs
to inform both teaching and policy making. Policy makers also need to
move away from controlling inputs and processes and instead increase the
emphasis on incentives to improve student learning, whether in school or
in higher education. And, finally, for these reforms to be financially fea-
Foreword xvii
sible, current spending on education needs to be used much more effec-

tively. In particular, countries in the region cannot afford to maintain one
of the lowest class sizes in the world while heating and lighting half-
empty buildings when resources are needed elsewhere.
It is our sincere hope that this book will stimulate debate about how
to improve education and training both for the countries under discussion
and in the world at large, and thereby encourage action to help realize
prosperity for all.
Philippe Le Houerou
Vice President
Europe and Central Asia Region
World Bank
Washington, D.C.
Acknowledgments
This study was managed by Lars Sondergaard, who authored chapters 3

and 4, and Mamta Murthi, who authored the Overview and chapter 7.
Jan Rutkowski authored chapter 1; Dina Abu-Ghaida authored chapter 2
with inputs from and Christian Bodewig and Lars Sondergaard; Lars
Sondergaard authored chapter 4 with inputs from Alex Usher; and
Christian Bodewig authored chapter 6 with inputs from Sarojini
Hirshleifer. Dina Abu-Ghaida co-managed the study in its early stages
with Lars Sondergaard.
The book grew out of a regional research program launched and sup-
ported by Europe and Central Asia’s Chief Economist’s office, and during
the writing of the book, two chief economists helped shape it: Pradeep
Mitra and Indermit Gill. The authors are grateful to both for their guid-
ance and support in writing this book.
The book is underpinned by significant contributions, including back-
ground papers, by Michael Mertaugh, Alex Usher, Iveta Silova, Andras
Benedek, Algerlynn Gill, Rostislav Kapelyushnikov, and Hakan Ercan,
and one jointly by Jerzy Wiśniewski, Maciej Jakubowski, Harry Anthony
Patrinos, and Emilio Ernesto Porta. We also thank Juan Manuel Moreno
and Nina Arnhold for reading and providing comments on various drafts
xix
xx Acknowledgments
and helping to shape the book’s main messages, and Algerlynn Gill for
her excellent research assistance and inputs into several of the book’s
chapters. In addition, Nadezhda Lepeshko and Carmen Laurente helped
prepare graphs and figures for the document. The book also benefitted
from the excellent editorial work of Peggy McInerny and Patricia Carley.
Larry Forgy provided editorial assistance on chapter 1.
We also benefitted from valuable comments from our peer reviewers:
Richard Murnane, Amit Dar, Bernard Hugonnier, Halsey Rogers, and
Manfred Wallenborn. And we benefitted from inputs, comments, and
suggestions from numerous colleagues, including Mohamed Ihsan
Ajwad, Gordon Betcherman, Mary Canning, Isak Froumin, Sachiko
Kataoka, Arvo Kuddo, Toby Linden, Lily Mulatu, Bojana Naceva,
Reehana Rifat Raza, Alberto Rodriguez, Marcelo Selowski, Jan Sadlak,
Jamil Salmi, Luis Crouch, and countless others. Any and all errors that
remain in this volume are the sole responsibility of the authors.
Abbreviations
AES Adult Education Survey

AHELO Assessment of Higher Education Learning
Outcomes
ALG adult learning grant
ALMP active labor market policy
ARACIS Romanian Agency for Quality Assurance in
Higher Education
BEEPS Business Environment and Enterprise
Performance Survey
BGN Bulgarian currency (lev)
CHE Center for Higher Education Development
CIS Commonwealth of Independent States
CISSTAT Interstate Statistical Committee of the
Commonwealth of Independent States
CNEAA Romanian Council for Accreditation
CONVEyT (Mexico) National Council of Education for Life
and Work
Crostat Republic of Croatia Central Bureau of Statistics
CVETS Continuing Vocational Education and Training
Survey
xxi
xxii Abbreviations
EBRD European Bank for Reconstruction and

Development
EC European Commission
ECA Europe and Central Asia
ECTA European Credit Transfer and Accumulation
System
EHEA European Higher Education Area
EMIS Education Management and Information System
ENQA European Network of Quality Assurance
ESF European Social Fund
ESG European Standards and Guidelines
ESU European Students Union
EU European Union
EU10/10+1 EU10 comprises Bulgaria, the Czech Republic,
Estonia, Hungary, Latvia, Lithuania, Poland,
Romania, the Slovak Republic and Slovenia.
EU10+1 adds Croatia.
EU15 Austria, Belgium, Denmark, Finland, France,
Germany, Greece, Ireland, Italy, Luxembourg, the
Netherlands, Portugal, Spain, Sweden, and the
United Kingdom
EU27 Austria, Belgium, Bulgaria, Cyprus, the Czech
Republic, Denmark, Estonia, Finland, France,
Germany, Greece, Hungary, Ireland, Italy, Latvia,
Lithuania, Luxemburg, Malta, the Netherlands,
Poland, Portugal, Romania, the Slovak Republic,
Slovenia, Spain, Sweden, and the United
Kingdom
EUA European University Association
EURASHE European Association of Institutions of Higher
Learning
GDP gross domestic product
HEGESCO Higher Education as a Generator of Strategic
Competences Project
HRDF (Malaysia) Human Resource Development Fund
IEA International Association for the Evaluation of
Educational Achievements
ICT information and communication technology
ILA Individual Learning Account
ISCED International Standard Classification of Education
Abbreviations xxiii
ISCO-88 International Standard Classification of

Occupations 1988, International Labour
Organization
LFS Labor Force Survey
LITS Life in Transition Survey (jointly conducted by
EBRD and World Bank)
LSC (UK) Learning and Skills Council
LSMS Living Standards and Measurement Survey
MERI (Romania) Ministry of Education Research and
Innovation
MES Ministry of Education and Science
MOF Ministry of Finance
MTEF Medium-Term Expenditure Framework
NAVET (Bulgaria) National Agency for Vocational
Education and Training
NGO nongovernmental organization
NQF National Qualifications Framework
NVQ National Vocational Qualification System
OECD Organisation for Economic Co-operation and
Development
PB performance budgeting
PETS Public Expenditure Tracking Survey
PIAAC Program for International Assessment of Adult
Competencies, OECD
PISA Programme for International Student Assessment,
OECD
PIRLS Progress in International Reading Literacy Study
PPP purchasing power parity
SEE South Eastern Europe (for the purposes of this
report, Albania, Bosnia and Herzegovina, Croatia,
the former Yugoslav Republic of Macedonia,
Montenegro, and Serbia)
SIMCE Sistema de Medicion de la Calidad Education
SME small and medium enterprise
SSC (UK) Sector Skills Councils
TALIS Teaching and Learning International Survey
TIMSS Trends in International Mathematics and
Science Study
UNECE United Nations Commission for Europe
UNICEF United Nations Children’s Fund
xxiv Abbreviations
UNESCO United Nations Educational, Scientific and

Cultural Organization
UNESCO-CEPES Centre Européan pour l’Enseignement Supérieur
(European Centre for Higher Education),
UNESCO, Bucharest
VCOT Vermont Classroom Observation Tool
VET vocational education and training
Country abbreviations used in figures and tables:

ALB Albania
ARM Armenia
AUT Austria
AZE Azerbaijan
BEL Belgium
BGR Bulgaria
BIH Bosnia and Herzegovina
BLR Belarus
CYP Cyprus
CZE Czech Republic
DEU Germany
DNK Denmark
ESP Spain
EST Estonia
FIN Finland
FRA France
GBR United Kingdom
GEO Georgia
GRC Greece
HRV Croatia
HUN Hungary
IRL Ireland
ISL Iceland
ITA Italy
KAZ Kazakhstan
KGZ Kyrgyz Republic
LTU Lithuania
LUX Luxembourg
LVA Latvia
MDA Moldova
MKD Macedonia, former Yugoslav Republic of
Abbreviations xxv
MLT Malta
NLD the Netherlands
NOR Norway
POL Poland
PRT Portugal
ROU Romania
RUS Russian Federation
SRB Serbia
SVK the Slovak Republic
SVN Slovenia
SWE Sweden
TJK Tajikistan
TKM Turkmenistan
TUR Turkey
UKR Ukraine
UZB Uzbekistan
Note: These three-letter country codes are part of the International

Organization for Standardization (ISO) 3166 standard to represent
countries.
Overview
The countries of Europe and Central Asia (ECA)1 are currently emerging
from the deepest recession suffered by any developing region. Post-crisis
conditions are very different from those of preceding years. Financial
resources are more limited and more expensive, and export growth is
restrained by potentially slower growth in destination countries. Restoring
and sustaining growth in this context require reforms that boost com-
petitiveness and increase labor productivity. Such reforms are all the
more important given the shrinking of the working-age population in
many countries of the region.
Earlier boom years in the ECA region exposed significant bottlenecks
to growth, particularly with respect to the skills of the labor force (Mitra
et al. 2010). Paradoxically, for a region with relatively high and expanding
educational attainment (as measured by the number of years of com-
pleted schooling) and relatively high-quality education in the early years
of schooling, a shortage of worker skills has emerged as one of the most
important constraints to firm expansion. This book seeks to answer the
following questions: Why do ECA firms increasingly complain that they
cannot find graduates with the right skills? What can ECA countries do
to close the skills gap?
1
2 Skills, Not Just Diplomas
When trying to answer these questions, this book faces a fundamental

problem: data exist on the number of students who graduate (i.e., how
many diplomas are issued) in ECA countries, but internationally compa-
rable data on whether graduates of upper secondary and tertiary institu-
tions (from which the bulk of ECA graduates now enter the labor market)
have the right skills and competencies for the job market do not exist.
This book uses a range of different data sources to argue that the skills
problem in the ECA region relates more to the quality and relevance of
the education provided in ECA countries than to problems of access. A
central argument of the book is that ministries of education are con-
strained in a number of ways from effectively managing their education
and training sectors. The three most important and interrelated impedi-
ments to improving quality and relevance are the lack of systematic data
on key skills-related performance issues (i.e., how much students are
learning and whether they are finding jobs after they graduate), the legacy
of central planning, and inefficient use of resources.
Lack of data on student learning and employment outcomes makes it
difficult for education ministries to address the legacy of central planning,
which emphasizes centralized management based on inputs. Ministries of
education in the region continue to micromanage the sector using
detailed norms and regulations. This input-oriented style of management
leads to the inefficient use of resources and results in a rigid education
sector—not the type of flexible sector needed by ECA to create modern,
skilled workforces. This book highlights how these constraints manifest
themselves and then presents ways of overcoming them, relying on the
experience of ECA countries that have successfully addressed them,
together with international experience. Recommendations are presented
in separate chapters for pre-university, tertiary, and adult education.
The Skills Challenge

The ECA region had a well-regarded education system before the transi-
tion. While the intervening years have taken some of the shine off this
reputation, these countries continue to have strong achievements.
Notably, enrollments are high at all levels of education. The communist
legacy is particularly visible in low-income countries in the region, which
have the highest secondary enrollments in the world for their income
level. Tertiary enrollments, which have grown rapidly in the past two
decades, are also high relative to income levels, with the exception of a
few low-income Commonwealth of Independent States (CIS) countries,
such as Azerbaijan and Uzbekistan.
Overview 3
The ECA region is reasonably successful not only in terms of enroll-

ment rates, but also in terms of providing quality education in the early
grades. In particular, in international tests that measure student compe-
tencies in primary and lower secondary education, students in many (but
not all) ECA countries outperform their peers in countries with the same
income level (see figure O.1). For instance, students in Estonia, where
gross domestic product (GDP) per capita was $21,6442 in 2008, outper-
formed their peers in Denmark, Sweden, Iceland, and Austria—where per
capita income is almost twice as high—on the mathematics part of the
2009 Programme for International Student Assessment (PISA). The pic-
ture for educational quality within the ECA region is varied, however, as
more than several countries, such as Bulgaria, Romania, Montenegro,
Azerbaijan, and the Kyrgyz Republic, have students (in the early grades)
that underperform relative to their income level.
In spite of these positive achievements of ECA education systems, the
European Bank for Reconstruction and Development (EBRD)–World
Bank Business Environment and Enterprise Performance Surveys (BEEPS)
show that ECA firms’ perception of skills constraints changed dramati-
cally around 2005.3 By 2008, skilled labor shortages had become the
second most commonly reported constraint to growth in the BEEPS sur-
vey across all countries in the region, second only to tax rates (see figure
O.2). On average, 30 percent of firms considered education and skills to
Figure O.1 Analysis of Reading Competency of 15-Year-Old Students

on the PISA 2009
550
POL EST
2009 mean performance
500 HUN
LVA SVN
TUR HRV CZE
LTU SVK
450 RUS
SRB
BGR
ROM
MNE
400 KAZ
ALB
AZE
350
KGZ
300
0 20,000 40,000 60,000 80,000 100,000
2008 GDP per capita, PPP (current international $)
Source: PISA 2009 Database and World Bank staff calculations based on World Development Indicators Database.
Note: PISA = Programme for International Student Assessment. The figure shows a regression line representing
countries’ predicted PISA reading scores based solely on GDP per capita, compared to the Organisation for Eco-
nomic Co-operation and Development mean reading score (horizontal line) and GDP per capita in 2008. See
“Abbreviations” for a key to country abbreviations.
Figure O.2 Distribution of Firms in ECA Region that Consider Worker Skills a
“Major” or “Very Severe” Constraint, 2008
9
mean = 30.2
8
TJK
7
UZB
number of countries
6
AZE CZE
5 ARM
4 BIH GEO EST
KOS KGZ HRV UKR

3
ALB LVA MDA
2 MKD
MNE SRB BGR POL LTU KAZ
1
HUN SVN TUR SVK ROM RUS BLR
0
%
%
10
20
30
40
50
60
70
80
90
x<
x<
x<
x<
x<
x<
x<
x<
x<
≤
≤
%
%
10
20
30
40
50
60
70
80
EU10+1 and Turkey Western Balkans LI CIS MI CIS
Source: Authors’ calculations based on BEEPS 2008.

Note: LI = low-income, MI = middle-income, x = % of firms (in respective countries) that consider education
as an obstacle. Figure shows data obtained from the fourth round of the BEEPS carried out in 2008–09,
which covered approximately 11,800 enterprises in 29 countries. See “Abbreviations” for a key to country
abbreviations.
be a major or severe constraint in 2008. The highest proportion of firms

reporting constraints were found among the middle-income CIS coun-
tries, where upwards of 40 percent of firms were dissatisfied with the
availability of skilled workers. A smaller proportion of firms in the Western
Balkans reported similar levels of dissatisfaction, with significant variabil-
ity in the dissatisfaction levels of firms in the EU10+1 (European Union
10+1) countries and Turkey (BEEPS dataset 2008). While the recession
may have provided something of a respite from these labor shortages, as
the economies in the region recover, labor demand will tighten once again
and skilled labor shortages will likely be worse than before.
Why Are Skills an Emerging Problem if Education

Systems Are Delivering?
Despite generally high average enrollment and attainment rates, as well as
respectable quality education (where measured, at the early grade level)
Overview 5
for their income level, ECA countries’ growth is constrained by skills

shortages. This is a puzzle that cannot be fully resolved, mainly because
crucial information is missing. Put simply, there is no internationally com-
parable information on the quality of upper secondary or tertiary educa-
tion or the relevance of education at these levels—information that
would indicate whether graduates have the right skills for the modern job
market
There are international assessments that provide information on stu-
dent competencies up to the age of 15 (usually the end of lower second-
ary education), but this is not an age when most people in the region are
entering the job market. Beyond these early-stage assessments, no com-
prehensive, fully reliable information exists on student competencies.
What is more, the information on student competencies that is available
relates only to fundamental cognitive skills, not behavioral skills—
involving such issues as work ethic and teamwork—which are empha-
sized by employers. For policy makers to better understand the causes
of the emerging skills bottleneck—and how to address it—this informa-
tional gap needs to be closed. Given the lack of data on the skills of new
labor market entrants, the book’s conclusions are necessarily based only
on the data that are obtainable.
Two problems related to quality seem particularly acute in the ECA
region: too many students are failing, and education systems have dif-
ficulties imparting problem-solving skills. In terms of the many stu-
dents that fail, PISA 2009 data show that all ECA countries (except a
handful of new EU member states) have very large proportions of
15-year-olds who have such poor numeracy and literacy skills that their
success in the modern workplace is highly doubtful. As work becomes
more demanding even for this group, their poor skills will become more
apparent. Although ECA countries perform strongly on international
assessments that measure students at the primary level (grade 4), their
performance is weaker on assessments of students around the end of
lower secondary education (grade 8). These results suggest that their
education systems are adept at imparting basic skills, but have problems
imparting higher order skills such as problem solving—the very skills
that firms increasingly seek.
More troubling, the quality of education in the ECA region also does
not show signs of consistent improvement. In fact, up to 2006 it appeared
to be getting worse in many countries. Rather than narrowing the gap
between their scores and Organisation for Economic Co-operation and
Development (OECD) mean scores, many ECA countries that participate
in international learning assessments at this level—which are likely the

better-managed countries—either seemed stuck or had regressed in terms
of their scores. It is probable, moreover, that the situation is even worse in
ECA countries that do not participate in efforts to measure and compare
the academic performance of their students. Encouragingly, the latest
round of PISA (2009) may have marked a turning point for 11 ECA coun-
tries that showed improvements compared to their 2006 performance
(including significant increases in Bulgaria, Romania, Serbia, the Kyrgyz
Republic, and Turkey). However, seven ECA countries continued the
downward slide observed previously with a deterioration compared to
their 2006 performance.
The quality of upper secondary and tertiary education in the region is
also unlikely to have improved, although lack of data makes it hard to
confirm this contention. Since educational quality is not showing consis-
tent improvement at the lower secondary level, this weakness is probably
mirrored at the upper secondary level (albeit with a lag). Enrollments at
the tertiary level used to be tightly controlled and reserved for the few,
best performers. In the past two decades, however, enrollments have
doubled, tripled, or quadrupled, but without the benefit of the quality
assurance mechanisms needed for the many new programs, institutions,
and types of students, and without the kind of information required for
parents and students to make informed choices. As a result, it is unclear if
the expansion of the system has been accompanied by growing quality.
Graduates of upper secondary and tertiary education may also be
graduating with the wrong set of skills. During the early years of the tran-
sition, the vocational school system—which once produced more than
half of all secondary graduates in most ECA countries—collapsed very
quickly. Voting with their feet, students have left vocational schools in
favor of general secondary education and the prospect of pursuing a ter-
tiary degree. It may be that the pendulum has swung too far in one direc-
tion, however, as employers assert that it is now hard to find graduates
with technical skills. Yet, until the region’s vocational schools are reformed
to ensure that they can produce graduates with appropriate skills, it may
be too soon to be encouraging students to return to this sector.
The contribution of education to the skills gap is not only a story of
the uncertain quality and relevance of formal education. It also involves
too little progress in providing workers with options for further training.
Growing evidence shows that continuous adult education and training
fosters employment and greater productivity. The development of this
sector should accordingly be central to the region’s economic growth
Overview 7
agenda, particularly in more advanced ECA countries and countries that

are facing a significant demographic decline—a decline that suggests that
labor force participation and individual productivity will have to increase
in order to generate higher rates of economic growth. Yet many of these
countries have only started to prepare for an expansion of the adult train-
ing sector, needed not only to compete in the global economy, but also to
address the lag effects of the transition to market economies. In light of
growing employer demands for greater skills, adult education and training
in ECA countries can no longer be ignored.
Priority Areas for Action

As noted at the outset, three major impediments are inhibiting the cre-
ation of flexible and responsive education systems in the ECA region.
First, these systems have been operating “in the dark” because they design
policies and take management decisions without systematically collecting
and analyzing data on the learning and employment outcomes of stu-
dents and graduates. Second, the legacy of central planning has kept the
management of school systems highly centralized, with central policy
makers intensely involved in operational details. While virtually all OECD
countries have embraced performance-oriented management in educa-
tion since the 1980s (see OECD 2004)—albeit at different speeds—most
ECA countries continue to use management practices that focus on com-
pliance with detailed regulations and financing schemes based on inputs,
not outputs. This means that most local education authorities and school
principals lack the autonomy and authority to make crucial management
decisions for their own institutions, including how much and what type
of vocational content students can choose from and how many teachers
to employ. As a result, the system is inflexible and does not respond either
to labor market needs or changes in student numbers.
Third, education systems in the ECA region use financial resources
highly inefficiently. Nowhere is this more apparent than in the pre-
university sector, where few countries have adjusted teacher staffing levels
in response to falling student numbers over the past 20 years. Consequently,
student-teacher ratios have fallen sharply (and per student costs have
risen)—more so than in any other region in the world (see figure O.3).
This implies that scarce resources are tied up in paying meager salaries to
too many staff members and heating half-empty buildings.
To be fair, these impediments affect ECA countries differently and
vary accordingly to the level of education. However, no ECA country
Figure O.3 Primary School Student-Teacher Ratios in ECA Compared to

Other Regions of the World, 1990–2008
29
27
student-teacher ratio (primary)
25
23
21
19
17
15
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
East Asia and Pacific Europe and Central Asia
Latin America and the Caribbean Middle East and North Africa
Sources: EdStats database, plus authors’ linear interpolations for missing years.
Note: Actual data are marked with dots. One of the problems with international data on student-teacher ratios
(in both the World Bank’s EdStats database and the other international education databases) is that it is unclear
whether or not data for a particular country are reported on a full-time equivalent basis.
has fully escaped the legacy of central planning.4 That system focused
on controlling inputs—that is, it checked whether local actors were in
compliance with detailed norms for all inputs. Education ministries
remained uninformed, however, about whether students acquired skills
and competencies.
In terms of how these impediments affect the different levels of educa-
tion, they are most clearly apparent at the pre-university level, which
absorbs two-thirds of total education funding in the ECA region. The
same three impediments affect tertiary education in a slightly different
way, as this sector has already undergone significant reform over the past
20 years, during which time it has doubled or tripled in size in many
countries (see chapter 5). Because most ECA countries do not yet have
adult education sectors in the modern sense, these impediments cannot
be fairly characterized as problems of this sector. However, the policy
recommendations offered in this book do apply to adult education and
offer ideas and principles for how this subsector could be developed and
managed in the region (see chapter 6).
Overview 9
Managing Education Systems for Results

To address the three impediments outlined above, countries in the ECA
region need to manage their education systems for results. First and fore-
most, they need to collect more information on “performance,” that is, on
what students are learning, what jobs they are finding, and what salaries
graduates are earning. In addition, ECA countries could benefit from fol-
lowing the lead of OECD countries and several developing countries and
replace input-oriented management of the education sector with perfor-
mance-oriented management. This change implies devolving more
autonomy to frontline providers and putting in place accountability
mechanisms—for instance, in the form of performance contracts and
performance-based budgeting—that emphasize performance and not
compliance with norms. With a greater availability of performance data
and with frontline providers empowered with greater decision-making
power, the preconditions will be in place for education systems in the
region to become both more responsive to labor market needs and more
efficient.
Focus Attention on Learning and Employment Outcomes

Countries in the ECA region would benefit from greater measurement
of student learning and employment outcomes, information that they
can then use to design better policies. There is significant consensus on
the key skills and competencies that education is expected to deliver, as
well as standardized tests to measure them.5 Many countries in the
region already participate in international assessments, but have moved
more slowly than other regions in generating useful information from
national assessments. Although all school systems in the region have
some kind of national assessment system, many are in the early stages of
development (see figure O.4) and there is very little evidence to date
that ECA countries are using these assessments to inform policy. Rather,
national assessments continue to be seen as a certification mechanism,
one that confirms a student has gained a mastery of a predetermined
curriculum, thus enabling a diploma to be issued or the student to prog-
ress to upper secondary or tertiary education. Student assessments are
rarely seen as an opportunity to identify the strengths and weaknesses of
an education system.
When students graduate, ministries of education in the ECA region do
not systematically collect, analyze, and disseminate information on their
employment outcomes. Such information is, however, critical for policy
Figure O.4 Status of Measuring and Using Data on Student Learning Outcomes in
the ECA Region, 2009
several years of
experience in
several years of measuring,
experience with analyzing, and
own assessments making use of
and regular learning results to
early piloting of participation in improve education
own assessment international policy
instruments, assessments
some
participation in Bulgaria
not started, very international
early stages, no, or assessments Hungary
rare participation Latvia
in international Lithuania
assessments Albania
Romania
Croatia
Armenia Serbia
Czech Republic
Azerbaijan Slovenia
Estonia
Belarus
Georgia
Bosnia and
Herzegovina Macedonia, FYR
Kazakhstan Moldova
Kosovo Montenegro
Kyrgyz Republic Poland
Tajikistan Russian Federation
Turkey Slovak Republic
Turkmenistan Ukraine
Uzbekistan
Sources: Authors’ assessments based on data from UNICEF (2007, table 2.1); the extent of countries’ participation
in PISA, Progress in International Literacy Study (PIRLS) and Trends in International Mathematics and Science
Study (TIMSS) since 1995; and inputs from World Bank country experts. The UNICEF data is drawn from “Table 2.1:
Status of Reforms of Assessment and Examination Systems, 2006,” regarding “Introduction of other school exams
or assessments (e.g. basic school)” and “Introduction of sample-based national assessment.” In this table, UNICEF
scores the progress of countries on a range of 0 to 4, with 0 representing “not planned or started” and 4 repre-
senting “operational.” In addition, the figure uses World Bank staff compilations on the number of international
assessments in which each country has participated.
makers and higher education institutions because it helps them detect

which programs and fields of study are in high demand among employers.
Moreover, this data can help students make better choices about which
university and field of study to pursue. Again, some member countries of
the OECD (including Hungary, Italy, and the Netherlands), together with
Romania, provide examples in this area for others to emulate.
Overview 11
Introduce Autonomy and Accountability Based on Results

Overcoming the legacy of central planning involves moving away from
detailed norms and instead holding actors accountable for performance
(or results). At the central level, this implies that policy makers in the
ECA region would relinquish certain duties and assume others so that
education systems may innovate and improve student learning outcomes.
The opportunity before central governments in these countries is to
move away from managing schools and classrooms and focus instead on
setting goals, policies, and standards; defining responsibilities; mobilizing
financial resources; ensuring political consensus; targeting poor and
excluded students; and monitoring and evaluating service delivery and
system quality.
Extending autonomy throughout education systems means placing
authority and responsibility in the hands of the people most able to
innovate and improve the quality of education: local managers and
education authorities (see Osborne and Gaebler 1992). It also means
holding these actors accountable for improving the learning out-
comes of the bulk of their students. Aligning the incentives of these
stakeholders with the student learning outcomes desired by policy
makers requires education ministries to set overall performance
goals; articulate who is responsible and accountable to whom and for
what; and ensure that these responsibilities are agreed, accepted, and
understood.
At the pre-university level, policy makers can expand the autonomy of
lower-level actors in the school system by granting them greater decision-
making power over school operations and budgets and relaxing norms on
class sizes. Greater autonomy is particularly needed in vocational educa-
tion and training, where programs and institutes need the ability to
expand or contract course offerings in response to student and employer
demand, not rigid governmental norms.
Simultaneously, policy makers will need to implement a range of
accountability mechanisms to hold schools accountable for results.
Potential mechanisms include the creation of school councils that
involve parents and local communities in school decision-making pro-
cesses; requiring schools to prepare school development plans that out-
line each school’s strengths and weaknesses, together with an action plan
for making improvements; preparing “school scorecards” that include
basic indicators on a school’s performance; as well sanctioning low-per-
forming schools and rewarding high-performing schools. Here, however,
it is important to define a high-performing school as one that delivers
improved learning outcomes to all types of students, including those

from low-income and minority households.
Significant improvements in the direction of greater autonomy and
accountability will, however, be difficult to achieve in the school system
unless policy makers in the region address the demoralization of the
teaching force, an effort that will involve resolution of serious overstaff-
ing and improving the pay, professional development, and work condi-
tions of teachers.
At the tertiary level, greater autonomy and effective accountability are
equally critical. The first part of this equation has been carried out in a
number of countries in the region—mainly the new EU member states
and, to a lesser degree, the countries of South Eastern Europe as well as
the Russian Federation and Ukraine—which have already granted tertiary
providers considerable autonomy. These countries, while continuing to
make needed improvements in the devolution of managerial authority,
must now face the primary (but not exclusive) challenge of improving
academic and fiscal integrity in the sector by introducing stronger
accountability mechanisms. Other countries in the region, especially
Belarus and most of the low-income CIS states, have yet to change how
their tertiary education systems are managed. These countries face the
simultaneous challenges of enhancing university autonomy while devel-
oping accountability mechanisms and introducing performance-based
financing. Regarding the need to monitor and evaluate system delivery
and quality, all countries in the region lag behind; in other words, all need
to make greater efforts to “turn on the lights” by vastly improving their
data collection on student learning and educational outcomes.
Even though fundamental integrity problems remain widespread in
the university sector, it is important that ECA policy makers refrain from
attempting to resolve these problems before they devolve authority to
universities and introduce flexible financing. The group of ECA countries
that have not yet initiated tertiary reforms need to introduce autonomy
and accountability as mutually reinforcing policy instruments. Policy
makers in these countries need not simply transfer autonomy to static,
traditional universities, but can instead mandate that increased autonomy
be accompanied by more “businesslike” and accountable leadership and
management.
In view of the relatively weak institutional capacity in the region—
including somewhat newly established or nonexistent quality assurance
agencies—policy makers may need to employ many different, comple-
mentary tools to strengthen accountability for results. These tools include
Overview 13
rankings, tracer studies, and creation of an enabling environment to

facilitate the growth of the private sector. Counting on a quality assurance
agency to quickly establish the capacity needed to externally monitor a
rapidly growing sector is risky—this kind of agency needs to be part of
the solution, but not the only response.
Improve the Efficiency of Resource Use through

Performance-based Financing
Managing school systems for performance means moving away from
inflexible line-item budgeting towards greater use of delegated budgets,
with incentives for maintaining enrollment and attendance (i.e., per stu-
dent financing—a path on which the majority of ECA countries have
already embarked). More flexible, smarter financing in the form of block
grants (i.e., contract- or performance-based) can also provide funding to
institutions in return for meeting agreed learning outcomes. Not only
does this type of financing give local education managers much-needed
flexibility that input-based budgets do not permit, it keeps them focused
on student results.
In addition to performance-based financing, improving financial effi-
ciency at the pre-university level will require creating incentives for larger
class sizes, encouraging school optimization, and addressing the issue of
overstaffing. At the tertiary level, several ECA countries (for example,
Poland and Romania) have already adopted per student financing, shift-
ing their focus away from inputs and toward results. The remainder of
countries in the region will need to introduce performance-based financ-
ing to improve the efficiency of spending in this sector.
If ECA countries want to simultaneously raise participation rates
and increase the quality of tertiary education, more private resources
will also be needed—the cost pressures facing higher education in the
region are simply too great. Mobilizing private resources will help
strengthen results-oriented outcomes in the tertiary sector in two fun-
damental ways: (1) by increasing competition, as more nimble and
innovative private providers help bring innovations to public providers,
and (2) by increasing students’ connection to the education process.
When students and parents pay tuition fees—whether to a public or a
private provider—they are generally more demanding about the quality
and relevance of the education provided. There is no single ideal level
of funding for higher education, and no single ideal mix of public and
private funding sources. Different countries will make different kinds
of trade-offs.
Build the Foundations of Adult Learning Systems

Unlike pre-university and university education, adult education systems
are largely nonexistent in most ECA countries. Building such systems will
require shifting away from government-defined programs towards a well-
regulated market of private and public providers that deliver training
services to both working and unemployed adults. Governments in the
region will, however, continue to play a role in education and training for
the unemployed, often by contracting private providers to deliver needed
services. Policies that address market failures in this sector are especially
important to lay the groundwork for effective adult education systems. In
general, successful systems require a high degree of coordination and part-
nership between government agencies and the private sector, as well as
giving the demand side of training—that is, businesses and individuals—a
strong voice in determining training policy. Once a solid adult learning
sector is established, governments can then “steer” it by monitoring data
on program quality, promoting autonomy in return for accountability, and
improving the efficiency of government financing in the sector.
Expanding adult learning is a priority for advanced economies in the
region that are facing a demographic decline. Their challenge is to
ensure that existing coordination mechanisms function well and that
regulation and financing are used to facilitate the emergence of a private
sector-oriented adult education and training market. The principal prior-
ity of less advanced economies in the region experiencing a demographic
decline (i.e., many of those in South Eastern Europe and the middle-
income CIS countries) is to introduce a policy framework for adult
learning and create the tools needed to implement this strategy (e.g.,
coordination mechanisms, plus initial steps toward regulation). For the
less advanced economies in the region that are not facing a demographic
decline (i.e., those of low-income CIS countries and Albania), it may be
more productive to limit efforts to establish policy frameworks and
coordination mechanisms for this sector. For many countries, participat-
ing in the OECD’s Programme for International Assessment of Adult
Competencies (PIAAC) would be an important first step in understand-
ing the current skills and competencies of their work forces.
Conclusion
It will take time to steer education systems in the ECA region away from
a focus on inputs toward a focus on delivering a quality education to the
Overview 15
majority of students. Waiting to begin this process will only cause coun-
tries to fall farther behind in delivering the skills their economies need to
compete in a global economy. Overall, much greater measurement of
student learning outcomes is needed at all levels of education if policy
makers are to understand the strengths and weaknesses of their respective
systems and design effective policies to improve them. If the role of cen-
tral governments is redefined to focus on strategic policy, system goals,
and regulation, central policy makers will be freed to use incentives to
manage lower-level actors to meet goals and then measure the results.
In schools, this process will involve greater use of per capita financing,
more autonomy, and greater accountability for learning outcomes. At the
level of higher education, strengthening national quality assurance mech-
anisms is needed to improve the accountability of education providers,
as is greater dissemination of information on learning and employment
outcomes (e.g., results from tracer studies, or university rankings) to
inform student choice. In addition, existing financing incentives will need
strengthening through the introduction of more performance-based bud-
geting and, where warranted, greater institutional autonomy. Although
adult learning systems are critically important for the region, their priority
in individual ECA countries will depend on economic and demographic
conditions and other, perhaps more pressing education challenges. All
countries in the region can, however, begin to develop a coherent strategy
for adult education and training.
Notes
1. The countries included in the Europe and Central Asia (ECA) region include
Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria,
Croatia, Czech Republic, Estonia, Georgia, Hungary, Kazakhstan, Kosovo,
Kyrgyz Republic, Latvia, Lithuania, the former Yugoslav Republic of
Macedonia, Moldova, Montenegro, Poland, Romania, the Russian Federation,
Serbia, the Slovak Republic, Slovenia, Tajikistan, Turkey, Turkmenistan,
Ukraine, and Uzbekistan.
2. Adjusted for Purchasing Power Parity, or PPP.
3. The EBRD–World Bank Business Environment and Enterprise Performance
Surveys (BEEPS) provide successive snapshots of the self-reported constraints
to expansion by firms in the ECA region. The surveys were conducted in
virtually all ECA countries in 1999, 2002, 2005, and 2008. Survey samples
are constructed by random sampling from the national registry of firms (or
equivalent) and cover both industry and service sectors.
4. It should be noted that Turkey does not have a socialist past, but does operate
“in the dark” to a large extent and has not made the performance-oriented
management reforms of other OECD countries. The expenditure efficiency
issues are, however, different in nature. See World Bank 2005a.
5. While the use of learning assessments is not uncontroversial, testing method-
ologies have improved in recent years, making these tests better and more
useful for assessing student learning than was previously the case.
CHAPTER 1
The Demand for Skills in ECA
Throughout the world there has been an increase in the demand for
skilled labor in what is now a global knowledge economy. More specifi-
cally, the demand for skills related to complex communication and non-
routine cognitive tasks has increased, while the demand for routine and
manual skills has declined. Employment trends in the region encompass-
ing Eastern Europe and Central Asia (ECA) reflect this new reality: jobs
have been shifting significantly from unskilled to skilled labor and the
wages of highly skilled workers have dramatically improved over the past
20 years. This shift parallels an economic transition that has involved
intensive enterprise restructuring as countries moved from centrally
planned to market-based economies. New skills have come into demand,
as employment has been allocated away from declining industries and
firms toward expanding ones. In addition, there has been a large-scale
shift of jobs from agriculture and, to a lesser extent, industry towards the
service sector.
This evolution in the labor market is at the heart of major changes in
employment outlook throughout the region. The movement toward
greater use of skilled labor has in turn produced relatively greater unem-
ployment for low-skilled workers. Wages of highly skilled white collar
workers have also grown at a much faster rate than those of less-skilled
17
manual workers. As a result, returns to education and professional

skills—that is, their value on the marketplace—have sharply increased
during the transition. The parallel upward movement of employment
and wages points to a shift in demand rather than supply as the main
cause of these trends.
Research indicates that this change in demand has not been adequately
met by an adjustment in the supply of skills, resulting in a kind of skills
“mismatch” throughout the region. In fact, many modern firms in ECA
countries view the lack of necessary skills among potential workers as a
major impediment to their operations and development. These needed
skills include not just the ability to apply knowledge and solve tasks, but
also the ability to work as part of a team—just one example of the type
of behavioral skills increasingly desired by employers.
This chapter describes the evolution and patterns of demand for
skills in the entire ECA region (see “Abbreviations” for a list of country
groupings used in this report. The notion of “labor demand” has differ-
ent meanings in different ECA subregions because of vast differences in
income level and the degree of formality of the labor market. While in
many of the new member states of the European Union (EU), wage
(-dependent) employment represents a large share of total employment,
in other countries (e.g., the Western Balkans or Commonwealth of
Independent States [CIS] members), it represents a lower fraction
(about 30–40 percent). Furthermore, a substantial part of wage employ-
ment in many subregions is in the public service sector. Accordingly, the
results presented in this chapter for wage workers refer to a considerably
larger part of some subregions than others.
Background: The Demand for Highly Skilled Labor

in the Global Knowledge Economy
There is no question that skilled labor is increasingly in demand in today’s
modern economies. Skills requiring expert, inventive thinking are sought
after and skills involving repetitive tasks are not. In the United States,
for example, the task content of the work done by the labor force has
changed greatly over the past 40 years. As shown in figure 1.1, the pro-
portion of the U.S. labor force employed in occupations that make inten-
sive use of nonroutine cognitive tasks—both interactive and analytic—has
increased substantially since 1960. In contrast, the percentage of the
labor force employed in occupations involving routine cognitive and
The Demand for Skills in ECA 19
Figure 1.1 Trends in Routine and Nonroutine Tasks of U.S. Labor Force,
1960–2002
mean task input in percentiles of 1960 task distribution
65
60
55
50
45
40
1960 1970 1980 1990 2000
nonroutine cognitive (analytical) nonroutine cognitive (managerial)
nonroutine manual routine cognitive
routine manual
Source: Autor, Levy, and Murnane 2003. With data updated to 2002 by David Autor.
manual tasks, as well as nonroutine manual activities, has declined (Autor,

Levy, and Murnane 2003).
There are three labor market ramifications of the shift in employment
from routine and manual tasks towards nonroutine cognitive tasks. First,
employment structure changes: a growing proportion of the labor force is
employed in jobs requiring higher-level, nonroutine skills (see Levy and
Murnane 2004). Second, wage structure changes: wage inequality rises
due to an increase in the skill premium; in particular, the wage gap
between low- and highly skilled workers widens. Third, unemployment
rises among less-skilled workers following the elimination of routine jobs.
Technological development also spurs change in how work is orga-
nized, which itself can lead to higher-skill needs. Computer technology,
for example, has contributed to the shift away from the so-called Taylorist
organization of labor (characterized by mass production and bureau-
cratic controls) to high-performance work practices (characterized by
decentralized decision making, just-in-time operations, teamwork, and

multitasking (EC 2008; OECD 1999b). U.S. data on firms that adopted
information and communication technology show, for example, that
increases in the demand for highly skilled workers can be attributed
more to the requirements of new work organization than to the introduc-
tion of new technology itself (Breshnan, Brynjolfsson, and Hitt 2002), a
process that has been termed “skill-biased organizational change.”1
Workers in today’s economies are increasingly required to solve com-
plex and unexpected tasks in their jobs, which involve fewer and fewer
simple, predictable activities. In addition, workers must be able to master
changing technologies and make sense of large amounts of information.
Figure 1.2 shows the complexities of this new working environment, as
expressed by worker responses to questions about the nature of their
employment.
Demand for Skilled Labor Has Risen in the ECA Region

This section focuses on two measurable aspects of skills in the ECA
region from a labor market perspective:2 educational attainment and
Figure 1.2 Worker Responses to Survey on Working Conditions in Europe, 2005
Does your main job involve: (% of respondents answering "yes")

90
80
70
60
percent
50
40
30
20
10
0
le en
th new
ta ng
s
w ge
k
as
ob se
of an
ex ki
s
gs
s
k
m
sk
or
st
pl rta
pr ore
ng
ds ch
in
ou
m e
ni
ho or
f
co und
un
on
ar
et e
le
ot
m oos
g
in
on
lv
ch
m
so
EU15 member states (i.e., non-ECA) new EU member states
Source: Authors’ calculations based on data obtained from Eurofound 2007.

occupation. The more educated a worker, the more skilled he or she is

assumed to be. It is taken for granted that some occupations (e.g., manager
or professional) require higher-level skills than others (e.g., service worker,
salesperson, machine operator). Education levels cannot be mapped
directly to occupations and skills, but education levels and occupations
can shed light on the demand for skills in a given labor market. The text
that follows relies on the International Standard Classification of
Occupations, known as ISCO-88 rev. 3, to classify various types of
employment in ECA countries.3
While data are not available that would allow the kind of analysis pre-
sented in figure 1.1 for the ECA region, a wide range of proximate data
points to similar trends there. Shifts in demand in favor of more skilled
labor have been reinforced in ECA by the transition from plan to market.
Turkey, although not a transition economy, has also experienced deep
structural changes associated with economic modernization and rapid
growth that have brought about changes in the demand for skills.
The economic transition in the region has been associated with
significant job flows between sectors, with labor moving to the service
sector, largely from agriculture and, to a lesser degree, from industry.
In reality, this transition simply removed institutional barriers to a natu-
ral restructuring process associated with economic development. For
example, in Estonia, service sector employment as a share of total
employment increased by 18 percentage points from 1990 to 2007; in
the Slovak Republic, by 12 percentage points from 1994 to 2007; in
Moldova, by 13 percentage points over the seven-year period, 2000 to
2007; and in the Kyrgyz Republic, by 5 percentage points in just four
years. In Turkey, again, not a transition economy, the share of services in
the national economy increased by 15 percentage points between 1990
and 2006. In all cases, increases in service employment as a share of total
employment were accompanied by some drop in the share of employ-
ment in industry, but mainly by a sharp drop in the share of employment
in agriculture.
There has also been significant job reallocation among enterprises
within the same industry in ECA countries (Rutkowski and Scarpetta
2005). In fact, job reallocation rates in transition economies exceed those
in developed countries (Bartelsman, Haltivanger, and Scarpetta 2004).
This reallocation can be attributed in some measure to firm turnover
(i.e., the entry and exit of firms), particularly in transition economies. For
example, in a number of ECA countries for which firm-level data are
available, newly established firms account for between 25 and 50 percent
of total job creation, exceeding the share common in developed countries

(Rutkowski and Scarpetta 2005).
Looking at employment structure by occupation provides a straight-
forward measure of the skill content of labor demand. In most ECA
countries, employment has shifted away from less skilled occupations
towards more skilled occupations. In particular, there has been a shift
from blue- to white-collar occupations, as well as an increase in the skill
content of employment in both kinds of work. The typical pattern of
change in occupational structure has included (1) a substantial fall in
demand for agricultural skills, (2) a fall in demand for manual labor, both
skilled and unskilled, (3) a rise in demand for service sector occupations,
and finally, (4) a surge in demand for professional skills.
Figure 1.3 shows the occupational profile (and, by implication, skill
profile) of newly created jobs versus jobs being destroyed in Georgia in
2007. As seen in the figure, a gradual but measured shift from industry
to services is occurring, with many jobs destroyed and few created for
such manual labor-oriented professions as technicians and machine
operators. Viewed in terms of labor demand, workers in these occupa-
tions saw the demand for their skills fall. At the same time, few jobs were
Figure 1.3 Job Creation and Destruction Rates for Selected Occupations in
Georgia, 2007
40
35
30
25
percent
20
15
10
5
0
s
ls
bl rs
ry
s
er
an
rk
le
pl
na
m to
ta
sa
cle
ag
s
ici
se era
en
io
er
pe
d
an
hn
ss
em
an
as p
ts
fe
m
d eo
af
te
el
o
ice
cr
pr
an hin
rv
se
ac
m
job creation job destruction
Source: Data obtained during a labor market survey of Georgian firms in 2007. The survey is discussed in greater
detail in Rutkowski 2008b.
destroyed and many were created for service workers and salespeople,
who thus saw the demand for their skills rise.
The pattern of job reallocation was somewhat different in Ukraine
(see figure 1.4). In 2007, demand shifted away from unskilled towards
skilled manual labor. The job destruction rate exceeded the job creation
rate for elementary occupations;4 consequently, the share of jobs requir-
ing elementary skills decreased. At the same time, the job creation rate
exceeded the job destruction rate for skilled manual occupations, mean-
ing the share of skilled manual jobs increased. There was also a decline in
professional employment in Ukraine.
Figure 1.5 shows patterns of change in the occupational structure of
employment across the ECA region over a longer time series, patterns
that are especially well illustrated by the panels for Poland (panel a) and
Russian Federation (panel d). Within the broad trend of the shift from
unskilled to skilled labor, there are numerous country-specific exceptions.
A diverse group of countries, from Romania to Turkey to the Kyrgyz
Republic, saw increases in the demand for manual labor (not decreases),
but these increases were for skilled manual labor. Some countries (e.g.,
the former Yugoslav Republic of Macedonia) saw a sharp increase in the
Figure 1.4 Job Creation and Destruction Rates for Selected Occupations in
Ukraine, 2007
45
40
35
30
percent
25
20
15
10
5
0
na d
ns
le
bl rs
s
rk
le
er
sio an
m to
op
cia
sa
cle
ls
r
s
bo
se era
er
es rs
pe
ni
d
of age
la
an
ch
as p
ts
d eo
af
te
ice
an
cr
an hin
m
rv
pr
se
ac
m
job creation job destruction
Sources: Ukraine Labor Demand Survey (World Bank 2009); Bank staff calculations.
Figure 1.5 Changes in Occupational Share of Total Employment in Selected ECA

Countries, Various Years
a. Poland, 1995–2006 b. Romania, 1995–2006
8
10
6
5
4
2 0
percent
percent
0 –5
–2
–10
–4
–6 –15
–8 –20
ag nd
ns
d ke s
sk rke al
sk rk ual
ke l
ag nd
ns
d ke s
sk rke ral
sk rker al
ke l
or ua
or ua
rk
ille wor sale
rk
ille wor sale

r
o u
cia
cia
w ultu
w ultu
an s a
an s a
cle
cle
te ers
ag rs
ille rs
d s
rs
te rs
ag rs
ille rs
rs
un wo n
w an
w an
w an
ille er
a
e
ni
ni
m nal
m nal
d
d
m
m
ric
ric
an
an
ch
ch
o
o
d
d
sio
sio
ille
ice
ice
es
es
rv
rv
of
of
un
se
se
pr
pr
sk
sk
c. Macedonia FYR, 2001–06 d. Russian Federation, 1995–2006
6 4
4 3
2 2
1
percent
percent
0
0
–2
–1
–4
–2
–6 –3
–8 –4
ag nd
ns
ke s
sk rke ral
sk rk ual
ke l
ag nd
ns
d ke s
sk rke ral
ille er l
ke l
or ua
sk ork ua
or ua
rk
or ale
rk
e
ille o al
cia
cia
w ultu
w ultu
an s a
an s a
cle
cle
te ers
rs
ille rs
d s
rs
te ers
ag rs
ille rs
d s
rs
un wo an
w an
un w an
w an
ille er
s
s
ni
ni
m nal
m nal
d
d
m
m
ric
ric
an
an
r
ch
ch
o
o
d
d
sio
sio
ag
w
w
ice
ice
es
es
d
rv
rv
ille
of
of
se
se
pr
pr
sk
sk
e. Kyrgyz Republic 2002–06 f. Turkey, 2001–06

10 10
8
6 5
4
2 0
percent
percent
0
–2
–5
–4
–6
–8 –10
–10
–12 –15
ag nd
ns
d rke es
ke l
sk rke al
ke l
ag nd
ns
d ker s
sk rke al
ke l
or ra
or ua
or ra
or ua
rk
rk
ille or sale
un wo anu
o u
ille wo sal
cia
cia
w ultu
w ultu
an s a
an s a
cle
cle
te rs
ag rs
ille rs
ille rs
rs
te ers
ille ers
ille rs
rs
w an
w n
w an
a
e
ni
ni
m nal
m nal
d
w d
m
m
k
ric
ric
an
an
ch
ch
d
d
io
io
ag
ice
ice
ss
ss
fe
fe
sk
sk
rv
rv
o
un
se
se
pr
pr
sk
sk
Sources: UNECE SD, Labour Force and Wages; Eurostat SD, Labour Market; World Bank staff calculations.
Note: The arrows in each panel suggest hypothetical movements of workers between occupations (i.e., possible
movements of labor between different occupations during this time period).
demand for medium-level professional skills (e.g., technicians and associ-

ate professionals), and a rise in unskilled manual employment. Others
(e.g., Georgia and Ukraine) saw a fall in the share of professional occupa-
tions. These variations notwithstanding, the overall picture is one of a
rising demand for skilled labor. The limited number of existing studies on
the demand for skills in transition economies also confirms this picture.5
Unemployment Patterns in ECA Countries Confirm the

Demand for Skilled Labor
Increased demand for skilled labor is also evident in the pattern of
unemployment and exit from unemployment found in the ECA region.6
Table 1.1 presents unemployment rates by educational attainment by
subregion.7 As a rule, the unemployment rate is highest among workers
with only a basic education, except in Turkey and the middle-income
CIS countries. The relatively low unemployment rate among workers
with this level of education can be attributed to the fact that many of
these workers are employed in agriculture.
By contrast, the unemployment rate is by far the lowest among work-
ers with a tertiary education in all ECA subregions. In fact, it is roughly
half that of workers with a general secondary education. In all subregions,
workers who finish school with vocational or technical skills are less
likely to be unemployed than workers with general secondary skills
(Turkey being an exception).8 The difference is particularly pronounced
in middle-income CIS countries, where the unemployment rate among
workers with a vocational or technical education is half that among work-
ers with a general secondary education.
Table 1.1 Unemployment Rates by Educational Attainment and

ECA Subregion, 2006
(percent)
Education level
Primary Secondary Secondary vocational
Country group or less general and/or technicala Tertiary
EU10 + 1 22.1 12.4 9.6 5.8
SEE 32.0 24.5 22.7 12.5
Middle-income CIS 7.2 14.2 7.2 5.0
Low-income CIS 29.7 26.1 22.3 16.2
Turkey 15.8 20.0 25.0 9.4
Sources: EBRD-World Bank 2006; authors’ calculations.
Note: See ”Abbreviations” for a key to country groupings. Unemployment rates estimated using data from the
EBRD-World Bank Life in Transition Survey (LITS) may differ from official country data because of methodological
differences.
a. The category of “secondary technical/vocational education” used by the LITS combines two different subcate-
gories: (1) workers who completed secondary technical school (4–5 years of courses leading to a secondary
school diploma), and (2) workers who received basic vocational training (1–3 years of courses, which do not lead
to a secondary school diploma). Evidence from countries for which the more detailed breakdown is available
indicates that the employment chances of workers with secondary technical education are significantly better
than that of workers with basic vocational training. For example, in Poland the unemployment rate among work-
ers with basic vocational training tends to be one-third higher than that among workers with secondary techni-
cal education. This may be due to the inferior quality of training received during the shorter courses, or to a sort-
ing process whereby students who are less able or come from a disadvantaged background tend to self-select
into schools that offer basic vocational training rather than the more demanding secondary technical schools.
The high demand for professional and technical skills in the ECA
region is also evident when one looks at the movement of workers from
unemployment to jobs. Professionals and technicians across countries in
the region have by far the highest chances of escaping unemployment
and finding a new job. Figure 1.6 depicts outflow rates from unemploy-
ment to employment for two ECA subregions: the EU10+1 (European
Union+1) countries and low-income CIS countries (the pattern prevail-
ing in middle-income CIS and South Eastern Europe (SEE) countries is
similar to that in the EU10+1 group).
In the EU10+1 countries, the escape rate from unemployment is
highest among professionals and technicians, followed by medium-
skilled white- and blue-collar workers. It is lowest among unskilled and
agricultural workers. In low-income CIS countries, the pattern is some-
what different: medium-skilled, blue-collar agricultural workers have
better chances of finding a job than do medium-skilled white-collar
workers. But workers in more skilled occupations nevertheless find jobs
more easily than those in less-skilled occupations.
The poor employment prospects of less-skilled workers in the region
are also evident from the unemployment-to-vacancy (U/V) ratio, which is
high for less-skilled workers and low for highly skilled workers. Data on
the U/V ratio for Croatia, the Czech Republic, and Ukraine illustrate the
point (see figure 1.7). Job opportunities for different occupations are quite
similar in these three countries: the number of unemployed professionals
and technicians (skill levels 3 and 4, respectively, in the ISCO-88 frame-
work) per vacancy is low and that of unemployed unskilled workers (skill
level 1) is high. Employment chances are also relatively good for skilled
manual workers, especially craft workers (skill level 2). But job prospects
are limited for white-collar workers with medium-level skills (e.g., clerks,
service workers, salespeople). The demand for white-collar labor with
medium-level skills is limited relative to supply, while the demand for
blue-collar labor with mid-level skills is high relative to supply.
Putting together the information on employment presented in the
previous section with the data on unemployment presented in this
section, it is possible to obtain an estimate of shortages and surpluses
in the economies of the ECA region, particularly by comparing the
relative shares of employment and unemployment of different educa-
tional groups. Figure 1.8 shows that in countries as diverse as Croatia,
Poland, Russia, Uzbekistan, and Tajikistan (listed in order of gross
national income [GNI] per capita), there is a shortage of workers with
professional or technical qualifications and a surplus of workers with
Figure 1.6 Yearly Outflow Rates from Employment to Jobs by Occupation in Two
Subregions of ECA, 2006
a. EU10 + 1
elementary labor
machine operators and
assemblers
craftspeople
agricultural workers
service and sales
clerks
technicians
professionals
60 70 80 90 100
percent
b. Low-income CIS
elementary labor
machine operators and
assemblers
craftspeople
agricultural workers
service and sales
clerks
technicians
professionals
60 65 70 75 80 85 90 95 100
percent
Sources: EBRD–World Bank 2006; World Bank staff calculations.
Note: The yearly outflow rate from unemployment to jobs is the number of workers who were employed during
the reference week, but unemployed one year earlier expressed as a percentage of all unemployed.
Figure 1.7 Unemployment-to-Vacancy Ratio by Occupation in Croatia, the Czech

Republic, and Ukraine, Various Years
a. Croatia, 2006
25
20
15
U/V ratio
10
0
managers
professionals
technicians
clerks
sales and
services
agricultural
workers
craftspeople
machine
operators and
assemblers
elementary
labor
- 4 3 2 2 2 2 2 1
ISCO skill level
b. Czech Republic, 2006

12
10
8
U/V ratio
0
managers
professionals
technicians
clerks
sales and
services
agricultural
workers
craftspeople
machine
operators and
assemblers
elementary
labor
- 4 3 2 2 2 2 2 1
ISCO skill level
c. Ukraine, 2007
10
9
8
7
6
U/V ratio
5
4
3
2
1
0
managers
professionals
technicians
clerks
personal
service
Sales
agricultural
workers
craftspeople
machine
operators and
assemblers
elementary
occupations
- 4 3 2 2 2 2 2 2 1
ISCO skill level
Sources: Data provided by the Croatian Employment Service for 2006, Ministry of Labour and Social Affairs of the
Czech Republic for 2006, and Ukraine State Employment Service for 2007; authors’ calculations.
Figure 1.8 Labor Shortages and Surpluses by Level of Educational Attainment,

Selected ECA Countries
a. Croatia, 2007 b. Poland, 2007
120
70 100
60
80
shortage/surplus index
50
40 60
30 40 shortage
20 shortage
20
10
0
0
–20
–10
surplus
–20 –40
surplus
–30 –60
y
in nal
ry
in l
al
ry
l
ra
in na
ica
es
ica
ar
er
a
a
ne
g
tra tio
tra atio
im
rti
rti
rl
hn
g
en
hn
in
yo
ge
te
te
pr
ca
ec
yg
ec
c
ar
vo
vo
ry
yt
yt
ar
im
da
ar
sic
sic
nd
ar
pr
n
nd
nd
ba
ba
co
co
co
co
se
se
se
se
c. Russian Federation, 2007 d. Bulgaria, 2007

150 250
200
100
150
50 shortage
100 shortage
0 50
0
–50
surplus –50 surplus
–100 –100
s
al
ge
al
y
na
ica
es
es
sit
ar
er
er
le
tio
rti
rl
rl
hn
er
en
en
ol
yo
yo
te
iv
ca
ec
lc
yg
yg
un
ar
ar
vo
yt
ca
ar
ar
im
im
ni
ar
nd
nd
pr
pr
ch
nd
co
co
te
co
se
se
se
e. Uzbekistan, 2005 f. Tajikistan, 2003

250
100
200
80
150 60
shortage
shortage
100 40
20
50
0
0
–20 surplus
surplus
–50 –40
ic
y
na l/
ge
al
na l/
ra
sit
ar
ar
ar
tio ca
tio ica
s
r
ne
ne
lle
ba
im
nd
rti
l
l
er
ca ni
n
co
ge
ge
te
iv
vo ech
ch
pr
co
un
te
ry
y
se
ca
yt
ar
a
ry
er
nd
nd
vo
ar
a
w
nd
nd
co
co
lo
co
co
se
se
se
se
Sources: Crostat 2007; Bulgaria 2007; Poland 2007; Russian Federation 2007; World Bank 2005b, 2003a; authors’
calculations.
Note: The skill shortage or surplus index is calculated as (e/u–1)*100, where e and u are respectively employment
and unemployment shares of a given occupational group.
basic skills. This pattern is common in most ECA countries, regardless

of their income level or progress in the transition towards a market
economy. In addition to the skills mismatch between educational cate-
gories, there is also a mismatch within educational categories in certain
countries. For example, a study of Croatia found that among workers
with a tertiary education, there was a shortage of engineers and a sur-
plus of lawyers and art designers (Rutkowski 2008a).
It is important to note that there continues to be excess demand for
skilled manual workers at the same time that there is an excess supply of
unskilled workers, medium-skilled nonmanual workers, and agricultural
workers. There is some evidence that the excess demand for skilled man-
ual workers is the result of a shrinking supply of workers in this category.
The relative shortage of skilled manual workers could have occurred
in three ways. First, as seen in figure 1.5, there has been a shift in occu-
pational structure away from skilled manual occupations in some coun-
tries (e.g., Poland, Russia, and FYR Macedonia). Given poor employment
prospects at home, some workers—especially younger workers—may
have decided to migrate to Western Europe, where employment pros-
pects and wages are better. There is substantial anecdotal evidence that
this has actually been the case for the populations of the EU10 countries
(especially after EU accession in 2004 allowed for greater labor mobil-
ity). Poland and Ukraine (the first a member of the EU, the second a
nonmember) provide good examples of the widespread outmigration of
skilled manual workers.9 Second, the supply of skilled manual workers
is likely to have declined due to a shift from vocational towards general
secondary education that has occurred in most ECA countries (Poland
is again a relevant example). Finally, older workers with vocational skills
may have decided to drop out of the labor force.
An important question is whether the apparent simultaneous excess
demand for some workers and excess supply of others is the result of
friction in labor market adjustments, or whether it has a structural com-
ponent. While structural unemployment is difficult to assess, there
appears to be significant long-term unemployment in the region concen-
trated among low-skilled workers, often leading to poverty and social
exclusion (Alam et al. 2005). Retraining is often ineffective because the
skill requirements of newly created jobs differ significantly from those
of the jobs that were destroyed.10 Some degree of friction in the labor
market is customary, but the skill profile of labor demand is changing
fairly rapidly in ECA countries due to both technological progress and
the adoption of new production processes. There are thus good reasons
to believe there is a structural component to the labor surpluses and

shortages observed in the region.
The ability of the labor market—and the education system—to
respond to this disequilibrium is a question of market signals and institu-
tions. Among potential issues are whether wages are sufficiently flexible
to adjust to changing demand and supply conditions, labor market infor-
mation on job prospects by type of education and occupation is available,
educational institutions can adjust program offerings to the needs of the
labor market, and employment services can effectively match workers to
available jobs. The next sections examine whether labor market signals
(via returns to education) reflect the underlying changes in labor demand
in the region and whether employers are able to attract workers with the
right skills.
Wages Have Risen for Skilled Labor

Changes in the demand for skills are certainly reflected in the movement
of wages. Given the supply of skills, an increase in demand for certain
skills manifests as an increase in the relative wages of workers who possess
these skills. In the majority of ECA countries, the returns to education
have sharply increased and are presently comparable to those observed in
developed market economies. However, as shown in figure 1.9, returns to
education are still relatively low in some of the less advanced, slower-
reforming transition countries. The evolution of the occupational wage
structure unambiguously points to an increase in wage premia for white-
collar skills, particularly professional and managerial skills.
Figure 1.9 reveals that ECA countries in the early 2000s could be
divided into three groups: a small group where returns to schooling
were high by international standards, a dominant group where returns
were moderate, and a small but nonnegligible group where returns were
low. There is a clear association between the degree of progress of
economic transition to a market economy and the level of returns to
schooling in these countries (Flabbi, Paternostro, and Tiongson 2007).
The high-returns group included advanced transition economies (e.g.,
Hungary and Poland), while the low-returns group included mainly less
advanced economies (e.g., Romania, Tajikistan, and Ukraine).11 As a rule,
countries where returns to education were high were those where
employment had already shifted towards more skilled occupations,
whereas countries where they were low were those where there had
been less of a shift. At that time, the rates of returns to education in
Figure 1.9 Returns to Schooling in Selected ECA Transition Countries,

Various Years
Hungary 2002 11.1

Poland 2000 10.6
Russian Federation 2008 9.0
Kazakhstan 2001 8.1
Moldova 2003 8.0
Latvia 2002 7.8
Belarus 2002 6.9
Bulgaria 2003 6.7
Czech Republic 2002 6.6
Slovak Republic 2002 6.1
Tajikistan 2003 4.9
Romania 2003 4.2
Ukraine 2003 4.0
Georgia 2002 0.3
0 2 4 6 8 10 12
percent
Sources: Patrinos 2008 (Ukraine); Kapelyushnikov 2008 (Russia) and Yemtsov, Cnobloch, and Mete 2006 (Belarus,
Bulgaria, Georgia, Hungary, Moldova, Poland, Romania, and Tajikistan).
Note: Average return per a year of additional education estimated based on a Mincerian wage regression of
the form: lnw = α + ρs + β0x + β1x2 + γz + ε; where w represents the wage of an individual, s denotes years of
schooling, x denotes labor market experience (years), z is a vector of other individual and job characteristics
(e.g., gender, industry, etc.), and ε is an error term, which inter alia captures unmeasured differences in innate
ability. The parameter ρ represents the return to one year of schooling.
most ECA economies were broadly comparable with those of developed

economies (about 7.5 percent), but substantially lower than those of
developing economies (where the rate is above 10 percent; Psacharopulos
and Patrinos 2004).
The increase in returns to schooling during the economic transition
in the region has been driven by a surge in returns to tertiary educa-
tion.12 Wage premia to tertiary education have increased dramatically,
while those to secondary education have remained roughly stable.13 This
is well illustrated by the example of Hungary (see figure 1.10), where
the wage premium to university education more than doubled in the
wake of the transition. At the same time, the premium to secondary
education (both general and technical) rose only slightly higher than it
Figure 1.10 Wage Premia by Level of Education in Hungary, 1986–2004
300
250
200
percent
150
100
50
0
1986 1989 1992 1995 1998 2001 2004
primary vocational secondary technical
secondary general college university
Sources: Campos and Jolliffe 2007; authors’ calculations.

Note: Wage premia are relative to uncompleted primary education.
was before the transition. Similar trends occurred in other EU10+1

countries (Rutkowski 1996a, 2001). Although time series data for other
ECA countries are not available, it seems fair to assume that the increase
in the wage premium to university education was a decisive factor in
other countries, too. The sharp increase in returns to tertiary education
is consistent with the growth in demand for highly professional skills
documented earlier in this chapter.
While in most ECA countries workers with a tertiary education earn
considerably higher salaries than workers with a secondary education,
returns to secondary education are limited. Figure 1.11 illustrates this
trend for Russia and Turkey. In Russia, an average worker with a univer-
sity education earns over twice as much as a worker with only primary
education and nearly 60 percent more than a worker with general sec-
ondary education.14 In turn, a worker with general upper-level secondary
education earns only 20 percent more than a worker with less education.
Differences in the wage premia by type of secondary school are virtually
negligible. A very similar pattern prevails in Turkey.
The shift in demand for skills during the economic transition produced
a new wage structure that is remarkably similar across ECA economies
(see figure 1.12). Using the average wage of workers in elementary occu-
pations as a reference, managers earn 3.5–4.5 times as much; profession-
als, 2.5–3.0 times more; technicians, about 2 times more; and skilled
Figure 1.11 Wage Premia by Education Level in the Russian Federation and Turkey
a. Russian Federation, 2007

120
100
80
percent
60
40
20
0
y
ge
ge
ica ry
ve l
ve l
le era
le era
sit
hn a
lle
le
ec nd
l)
l)
er
er en
er n
ol
co
w ge
iv
l/t eco
lc
pp g
un
(u ary
(lo ary
ca
m
s
ni
so
nd
nd
ch
co
co
na
te
se
se
tio
ca
vo
b. Turkey, 2006
120
100
80
percent
60
40
20
0
y
ry
y
ica
ra
ar
ar
a
ne
rti
nd
im
hn
ge
te
pr
co
ec
ry
se
yt
a
ar
er
nd
nd
w
co
lo
co
se
se
Sources: Kapelyushnikov 2008 (Russian Federation); Ercan 2008 (Turkey); authors’ calculations.
Figure 1.12 Premia by Occupation Relative to Elementary Occupations in Various

ECA Countries
a. Bulgaria, 2006 b. Czech Republic, 2007
400 300
350 250
300
200
250
percent
percent
200 150
150
100
100
50
50
0 0
s
ke d
le
bl rs
d rke d
cr ers l
le
bl rs
na
na
k a
er
an
rk
er
an
rk
or an
ille o an
or ur
m to
m to
op
op
cle
cle
ag
ag
sio
sio
rs
ag rs
s
ici
ici
w ult
se era
se era
er
er
s w ce
sk s w ice
pe
pe
an
an
n
n
es
es
ric
le rvi
ch
ch
as p
as p
ts
ts
le rv
m
m
of
of
d eo
d eo
af
af
sa se
sa se
te
te
pr
pr
cr
an hin
an hin
ac
ac
m
m
c. Russian Federation, 2007 d. Turkey, 2006
300
250 300
200 250
percent
150 200
percent
100 150
50 100
0 50
–50 0
bl rs
bl rs
s
l
ns
ns
d ke d
d ork and
na
na
le
le
s
s
cr ker al
cr ker al
er
er
m to
m to
rk
rk
n
op
op
cia
cia
or ur
or ur
ag
ag
ille o a
sio
sio
s
s
cle
cle
se era
se era
ric s
ric s
er
er
af s
af s
ag r
ag er
sk s w ice
ille w e
w ult
w ult
pe
pe
ni
ni
an
an
es
es
sk les rvic
as p
as p
ch
ch
rv
ts
ts
r
m
m
of
of
d eo
d eo
sa se
sa se
te
te
pr
pr
an hin
an hin
l e
ac
ac
m
Sources: Eurostat, Labor Market Statistics (database) (Bulgaria and Czech Republic); Ercan 2008 (Turkey),
Kapelyushnikov 2008 (Russia); World Bank staff calculations.
manual workers and clerks, about 1.5 times more. Service workers and
salespeople earn only slightly more—some 10 percent—than workers in
elementary occupations. While patterns across most ECA countries are
similar, the relative wage status of skilled manual workers is much better
in Russia than in the EU10 countries. Concomitantly, the relative wage
status of professionals and technicians is worse in Russia. Moreover,
there are no returns to clerical skills in Russia; clerks earn on average the
same wage as laborers. Thus the occupational wage structure in Russia
is biased towards skilled manual labor; as such, it is closer to that prevail-
ing under central planning. A similar pattern also prevails in Ukraine
(World Bank 2009b).15
In addition to the current variance in wages across job categories,
there is evidence that wage rates are diverging over time. In most ECA
countries for which data are available, workers in occupations requiring
highly skilled labor (e.g., managers, professionals, and technicians) saw
their relative wages increase, while workers in occupations requiring
medium- and lower-level skills (especially manual skills) saw their rela-
tive wages fall. Thus the wage differential between highly skilled white-
collar workers and less skilled white- and blue-collar workers has
substantially widened. Figure 1.13 illustrates this trend in Bulgaria and
Poland.16 The wage status of highly skilled white-collar workers in these
countries has substantially improved, while that of less skilled workers
(both white- and blue-collar) has deteriorated during the transition.
Lack of Needed Skills Is Impeding Enterprise Growth

The preceding sections documented the shift in labor demand in ECA
countries from less to more skilled labor within both blue- and white-
collar occupations. In addition, demand has shifted notably from blue- to
white-collar occupations. These changes, which have resulted in high
rates of return to investment in skills in the region, are the result of two
simultaneous changes: an increase in demand for skilled workers arising
from skill-biased technological and organizational change and (in the
transition economies) factors specific to the transition from a centrally
planned to a market economy.
Available evidence suggests that the supply of skills has adjusted fairly
slowly to the changes in demand. Many firms in ECA, especially modern
and growing firms, now see the skills of available workers as a major
obstacle to their operations and growth. The strongest evidence for this
perception comes from the findings of the Business Environment and
Enterprise Performance Surveys (BEEPS) conducted by the European
Bank for Reconstruction and Development (EBRD) and the World
Bank.17 For the first time since BEEPS was launched in 1999, workers’
education and skills were recently identified by employers as one of the
top constraints to firm growth.18
In fact, worker skills were not typically identified by firms as a con-
straint until 2005; other constraints (e.g., taxes and business environ-
ment) were deemed more important. However, during the most recent
round of BEEPS, conducted in 2008, lack of skilled manpower moved up
to become the second-most common constraint on growth cited by firms.
Indeed, by 2008 the traditional strengths of ECA countries—good
endowments of infrastructure and skills—appear to have eroded (Mitra,
Selowsky, and Zalduendo 2010).
Panel a of figure 1.14 shows the distribution of ECA countries accord-
ing to the level of dissatisfaction with workers’ education and skills
among local firms. On average, 30 percent of firms considered these
Figure 1.13 Wage Growth by Occupation in Bulgaria and Poland, 1996–2006

(average wage growth = 1)
a. Bulgaria
1.6
1.5
1.4
1.3
wage growth
1.2
1.1
1
0.9
0.8
0.7
1996 2006
managers and officials professionals technicians
clerks service and sales craftspeople
machine operators and assemblers elementary occupations
b. Poland
1.3
1.2
1.1
wage growth
0.9
0.8
0.7
1996 2006
managers and officials professionals
technicians clerks
service and sales agricultural workers
craftspeople machine operators and assemblers
elementary occupations
Sources: Eurostat, Labor Market Statistics (database); World Bank staff calculations.
Figure 1.14 Employers’ Perceptions of Worker Skills as a Constraint to Growth,

Various ECA Countries
a. Distribution of firms that consider worker

skills a “major” or “very severe” constraint, 2008
9
mean = 30.2
8
TJK
7
UZB
6
AZE ARM CZE
countries
5
BIH GEO EST
4
KOS KGZ HRV UKR
3
MKD ALB LVA MDA
2
1
0
%
%
10
20
30
40
50
60
70
80
90
x<
x<
x<
x<
x<
x<
x<
x<
x<
<
<
<
<
<
<
<
<
%
%
%
%
60
80
10
20
30
40
50
70
b. Change in firms’ satisfaction with

worker skills between 2005 and 2008
10
9
8
7
6
frequency
5
4
3
2
1
0
5
.4
.3
.2
.1
x
.
0.
0.
0.
0.
0.
–0
≤
–0
–0
–0
–0
x<
5
x<
x<
x<
x<
x<
x<
x<
x<
x<
x<
0.
≤
≤
1
≤
.
–0
4
.5
.4
.3
.2
0.
0.
0.
0.
–0
–0
–0
–0
LI CIS MI CIS Western Balkans EU10+1 and Turkey
Sources: BEEPS 2008; Mitra, Selowsky, and Zalduendo 2010.

Note: LI = low-income, MI = middle-income. See “Abbreviations” for country abbreviations and groupings.
factors to be a major or severe constraint to their operations and growth.

The greatest dissatisfaction was among middle-income CIS countries,
where more than 40 percent of firms were dissatisfied. In contrast, rela-
tively less dissatisfaction was reported by firms in the Western
Balkan countries and significant variability was reported in the EU10+1
countries.
Worker skills are particularly likely to become a binding constraint
during periods of rapid economic growth. As might then be expected, the
percentage of firms for which the availability of the “right” workers was
either no obstacle or a minor obstacle decreased during the period between
2005 and 2008 due to strong economic growth in the region. Panel b of
figure 1.14 depicts the difference in percentage of satisfied firms between
2005 and 2008 regarding the suitability of the skills and education of
workers for doing business. A negative value (along the x axis) assigned
to a particular country indicates a decline in the percentage of satisfied
firms over this period. For example, if the share of satisfied firms in coun-
try A declined by 25 percent between 2005 and 2008, then country A
would be positioned in the –0.2 to –0.3 percent range of the x-axis. As
this panel shows, all ECA countries, with the exception of one EU mem-
ber state, experienced a decline in the share of satisfied firms between
2005 and 2008. The worst deterioration in the availability of skills rela-
tive to demand took place in the middle-income CIS countries. But even
in a large number of new EU members and other CIS countries, the
reduction in satisfied firms was between 0.2 and 0.3 (or 20–30 percent
of all firms).
As might be expected, modern firms suffer more from skills shortages
than do traditional ones.19 Across the ECA region, modern companies
find it more difficult to recruit workers with the required qualifications
than do traditional firms, particularly professional and skilled manual
workers. For example, in the EU10+1 countries in 2005, modern firms
needed, on average, five weeks to hire a professional worker, while tradi-
tional firms needed fewer than three weeks (see panel a of figure 1.15).
In low-income CIS countries, the search for a skilled manual worker
among modern firms was, on average, one week longer than that of tradi-
tional enterprises (see panel b of figure 1.15).
Skills Mismatch in the ECA Region

As the previous sections have documented, there is a skills mismatch in
the ECA region. Unless it is addressed, governments in the region will be
Figure 1.15 Time Needed by Firms to Hire Specific Workers, by

ECA Subregion, 2005
a. Modern and traditional firms seeking professional worker, 2005

6
4
weeks
0
+1
ey
A
SE
CI
CI
EC
rk
10
Tu
m
m
EU
co
co
in
-in
e-
w
dl
lo
id
m
b. Modern and traditional firms seeking skilled manual worker, 2005

4.5
4.0
3.5
3.0
weeks
2.5
2.0
1.5
1.0
0.5
0
+1
ey
A
SE
CI
CI
EC
rk
10
Tu
om
m
EU
co
c
in
-in
e-
w
dl
lo
id
m
modern firms traditional firms
Sources: EBRD-World Bank 2005; World Bank staff calculations.

Note: See “Abbreviations” for country abbreviations and groupings.
unable to facilitate either the growth of firms or a reduction in structural

unemployment. Why the skills mismatch exists is a subject of intense
inquiry.20 Although evidence indicates that many factors are involved, this
book focuses on one possible cause: education and training institutions in
the region provide inappropriate types of skills to their graduates.
There are several ways in which the skills mismatch can be related to
the education and training of graduates in ECA countries. First, gradu-
ates may not have the necessary technical skills to do the jobs that
employers have to offer. This is not solely a matter of educational attain-
ment, that is, of completing a certain level of education. It is also a mat-
ter of whether students have the skills needed on the job—the relevant
knowledge, the ability to apply that knowledge, and the know-how to
complete tasks and solve problems—when they graduate. In addition,
the relevance of the education received by graduates is also important.
That is, students may be studying the wrong things. Second, graduates of
education and training systems in the region may lack the necessary
behavioral (or soft) skills needed by employers, such as job attitudes and
teamwork skills. There is widespread evidence from across the ECA
region that firms seek graduates who not only have knowledge as well
as technical and general skills, but who also have behavioral skills.
For instance, recent surveys of employers in Kazakhstan (panel a of
figure 1.16) and Poland (panel b of figure 1.16) show that firms value
such behavioral skills as highly (in some instances, more highly) as
knowledge and routine cognitive skills (e.g., mathematics and computer
skills)—important as the latter may be.21
Before turning to the ways in which education and training systems
fall short of delivering the right skills, it is worth noting that one possible
reason for the continued identification of skills as a bottleneck to growth
in the ECA region is unreasonable expectations on the part of ECA firms.
A recent comparison of firms’ expectations of new workers in the United
Kingdom and Poland, for example, showed that Polish employers
expected young workers to possess advanced vocational job-specific
skills, while their U.K. counterparts preferred workers who had reached
an adequate level of literacy and numeracy, without necessarily having
fully developed, job-specific vocational skills (World Bank 2010b). The
report concluded that: “The expectations of Polish employers will prob-
ably have to change, and they will have to assume a greater role in the
financing and provision of vocational training to their workforce” (World
Bank 2010b).
Summary
The demand for highly skilled labor has increased in most if not all ECA
countries. This profound change in demand stems from the opening
of their economies to international competition and the consequent
% of firms reporting % of firms reporting
42
0
10
20
30
40
50
60
70
80
90
100
0
10
20
30
40
50
60
70
80
advanced
literacy
technical/vocational
and Poland
basic Russian language
technical/vocational
literacy analytical ability
problem solving
problem solving
they were categorized as “traditional.”

use of information customer
and communication relations skills
technology
general and technical skills
Sources: Ivaschenko 2008; World Bank 2010b.

writing skills
numeracy
mathematics
general and technical skills

foreign language (calculation)
innovative firms
computer skills
(percentage of firms reporting as “very important”)
responsibility
and reliability
b. Poland, 2009
Kazakh language
motivation and
a. Kazakhstan, 2008
commitment
English language
team-work ability
traditional firms
ability to work
independently
communication
soft skills
initiative and time management
ability to work
independently communication
planning and
soft skills
organizing telephone
conversation
customer care
management
skills
Figure 1.16 Employers’ Valuation of Workers’ Knowledge and Skills in Kazakhstan
Firms were categorized as “innovative” if in the year preceding the survey they introduced innovations, otherwise
far-reaching reallocation of labor that took place. New firms have entered
the market, new economic activities have emerged, and new jobs have
been created, calling for different skills than those required previously. As
economies have shifted, the service sector has expanded, while the agri-
cultural and, to a lesser degree, industrial sectors have shrunk. Firms have
changed the skills composition of their workforces to raise productivity
and improve competitiveness; as a result, labor demand has shifted away
from lower-skilled toward higher-skilled occupations.
This change in employment structure is most pronounced in the
countries of Central Europe, but is visible in some form everywhere in
the ECA region. Simultaneously, returns to education have increased and
the wages of highly skilled professional workers have greatly improved,
while those of lesser-skilled workers have deteriorated. Consequently,
the job prospects of better-educated, highly skilled workers have
expanded dramatically. Unemployment and low-paid employment in
the region is accordingly concentrated among less educated and less
skilled workers, a group whose job opportunities are increasingly limited
in most ECA countries.
All of these changes have put added strains on education and training
systems in the region, which are expected to prepare workers for changed
and changing labor markets, but which do not appear geared to meet
those labor market needs. The next chapter looks at the supply of skills
to the labor market in ECA countries in more detail. Given that skills
shortages have become a major constraint to firm growth in many ECA
countries, whatever changes have occurred on the supply side appear to
have been insufficient. The next chapter tries to explain why.
Notes
1. Alternatively, upgrading the education of the workforce may itself be a
factor that leads to increased demand for higher-level skills. That is, the
increase in demand for such skills may reflect an increase in the supply of
these skills, a process known as endogenous skill-biased technological
change. Acemoğlu (1998) argues that a high proportion of skilled workers
in the labor force implies a large market size for skill-complementary tech-
nologies, meaning that the productivity of skilled workers will be upgraded
faster. In this model, an increase in the supply of skills over the long run
induces skill-biased technological change, which shifts the relative demand
for skills (EC 2008).
2. Chapter 2 discusses skills from an education and training perspective.
3. In this book, high-level white-collar (i.e., nonmanual) skills refer to ISCO-88

groups 1 and 2, and high-level order blue-collar (manual) skills refer to
ISCO-88 groups 6, 7, and 8. Low-level, nonmanual skills refer to ISCO-88
groups 4 and 5, and low-level manual skills refer to ISCO-88 group 9. See
ILO (1988).
4. Elementary occupations consist of simple and routine tasks that mainly
require the use of hand-held tools and often some physical effort.
5. Peter (2003) analyzes skill-biased changes in wages and employment in the
Russian economy during the transition. Commander and Köllö (2004) exam-
ine the evolution of demand for skills in Hungary, Romania, and Russia.
Bartlett (2007) examines the demand for skills in the Western Balkans, and
Rutkowski (2008b) and the World Bank (2009b) document the skills mis-
match in Georgia and Ukraine, respectively.
6. At the beginning of the transition, the rising unemployment rates of low-
skilled workers likely in part reflected the hidden unemployment of these
workers—that is, workers whom the state claimed to be employed but who
were in fact superfluous to production. With the transition to a more market-
based economy, this de facto unemployment was essentially unmasked, rather
than created. However, by 2011, the rising unemployment rates of low-skilled
workers is more likely a reflection of actual shifts in demand for labor rather
than a continued unmasking of hidden unemployment.
7. Unemployment statistics by educational attainment can be difficult to inter-
pret because the determination of whether a respondent is in the labor force
affects both the numerator and the denominator of the unemployment rate.
If a larger share of workers with vocational secondary education are judged to
be “discouraged workers” and out of the work force, this may lead to a lower
unemployment rate for workers in this category, even if a smaller share of
them are employed. The more appropriate statistic to use for judging the
match between terminal level of education and employment status is the
employment rate (employed share of graduates with the specific education
level) and not the unemployment rate. Such calculations can be done if coun-
tries start tracking their graduates using “graduate tracer studies,” as recom-
mended in this book.
8. It can be argued that the better employment outcomes of workers with
technical secondary education compared with those of workers with gen-
eral secondary education reflect the different composition of both groups.
One hypothesis is that due to the selection process, graduates of general
secondary schools who do not proceed to tertiary education are less able,
on average, than the graduates of secondary technical schools. However,
this hypothesis remains to be tested. Another possibility is that workers
with technical secondary education are in a different age cohort—one
that received a better education—than workers with general secondary
education. Yet available evidence indicates that the cohort effect is small
and thus unlikely to account for the differences between the two educa-
tional groups.
9. While not under the scope of this book, it should be noted that migration has
had a significant effect on labor trends in the region. A recent World Bank
study examines both migration and remittances in the region. See Mansoor
and Quillin 2007.
10. Surveys of demand for skills that were carried out in Poland and FYR
Macedonia in 2009 indicated that most newly created jobs require profes-
sional or nonmanual skills, while most jobs that were destroyed required
manual skills. This obviously limits the scope for retraining (Rutkowski 2010;
World Bank 2010b).
11. The low returns to schooling for these countries is presumably in part a legacy
of the socialist period, when education levels were much higher than would
be predicted for their level of development, a situation that holds down wages
during the time when the economic structure is catching up to education
levels.
12. As figure 1.9 suggests, much of the increase in these returns was concentrated
in the early years of the transition. A few countries, typically those with lower
returns, show a flatter trajectory.
13. The terms “premia” and “returns” to education are used here interchangeably.
However, strictly speaking, they are not synonymous. Returns to education
take into account both wage premia and private costs of obtaining education.
Accordingly, premia could have increased without private returns increasing
as dramatically.
14. These estimates do not control for other factors that affect wages, such as
labor market experience, gender, location, and so forth.
15. This finding likely reflects the legacy of central planning and a noncompeti-
tive wage setting, especially in enterprises fully or partly owned by the state,
in both countries.
16. In some countries in the region there were deviations from the overall trend
presented here. For example, salaries of professionals in the Czech Republic
decreased slightly in relative terms between 1996 and 2007. Similarly, in
Estonia and the Slovak Republic, the relative wages of managers fell slightly
during the same period. The slight fall in the relative wages of professionals
and managers that occurred in the mid-2000s likely represents an adjustment
to the increase in the supply of the relevant skills that occurred in response
to growing demand.
17. BEEPS, a joint initiative of the EBRD and the World Bank, collects firm-level
data on a broad range of issues concerning the business environment and the
performance of firms, including labor and workforce skills.
18. The most common constraint identified by firms in the region was the tax
regime.
19. Firms were categorized as “modern” if they used the Internet in interactions
with their clients and suppliers, and “traditional” if they did not. Clearly, use
of the Internet is only a proxy for modern technology.
20. Some degree of skills mismatch is natural and unavoidable in every growing
and restructuring economy due to adjustment costs and the delayed response
of national educational systems. Especially in the context of fast-changing
technology, changes in the skill structure of the labor supply tend to lag
behind those of labor demand, even in a well-performing labor market.
21. Surveys of the demand for skills in Poland and FYR Macedonia that were car-
ried out in 2009 provide additional empirical support to the claim that
employers value job attitudes and behavioral skills at least as much as aca-
demic or technical and/or vocational skills. At the same time, both Polish and
Macedonian employers complain that many recent school graduates and
young workers are lacking “soft” skills (Rutkowski 2010; World Bank 2010b).
CHAPTER 2
Education and the Supply

of Skills to the ECA Market
The skills gap described in the previous chapter represents something of

a puzzle from the perspective of the education and training sector in the
ECA region. ECA countries have had very high educational attainment
rates for decades and, over the past 20 years, have experienced a rapid
expansion of tertiary enrollment. In addition, these countries deliver
quality education (where it is measured) at the elementary and lower
secondary levels comparable to that of their peers at the same income
level. Given these conditions, why do ECA firms increasingly complain
that they cannot find graduates with the right skills?
In answering this question, there is a fundamental problem: data are
available on the number of students who graduate (i.e., how many diplo-
mas are issued) in ECA countries, but internationally comparable data on
the quality and relevance of upper secondary and tertiary institutions do
not exist. Data gaps are even more acute with regard to the education and
training received by adults outside of the formal education system. Only
a few of the new EU member states track the extent of such training (in
terms of participation rates and hours of training); data on the quality and
relevance of this type of training, however, do not exist anywhere.
While the data gaps make it impossible to provide a definitive answer
to the skills puzzle, this study is able to point to fundamental weaknesses
47
in the ECA region’s education systems that almost certainly contribute to

the skills gaps observed in chapter 1. Addressing these shortcomings may
not fully close these gaps, but it will create more flexible and better-
managed education systems, which will, in turn, increase the likelihood
that young people will graduate with—and adult workers will be able to
acquire—the skills in demand today. Moreover, it is clearly a priority that
ECA countries address the information deficits that are currently forestall-
ing a more comprehensive—and satisfying—answer to the puzzle above.
Although the education offered in many ECA countries is respectable
for their level of income, it does not appear to be good enough (or of the
right relevance) to meet the rising demand for skills in the region. Two
problems related to quality seem particularly acute: too many students are
failing and education systems have difficulties imparting problem-solving
skills. Educational quality, moreover, is not demonstrating reliable progress.
It is not improving at the lower secondary level, a weakness that is proba-
bly mirrored at the upper secondary level (albeit with a lag). Enrollments
at the tertiary level have doubled, tripled, or quadrupled in the past two
decades, but this occurred in an environment where quality assurance
mechanisms are very weak, indicating that quality in this sector is also
unlikely to have improved. The skills gap can also be explained by too little
progress in providing workers with options for continuous training, despite
evidence that such training fosters employment and greater productivity.
The development of the adult training sector should accordingly be central
to the region’s medium-term economic growth agenda.
Background: The Global Knowledge Economy Requires

Lifelong Learning
To prepare students for the global knowledge economy, education sys-
tems must build human capital, which the OECD defines as “the knowl-
edge, skills, competencies, and other attributes embodied in individuals
that are relevant to personal, social, and economic well-being” (OECD
1999a). The framework for education is thus one of lifelong learning in
which the new knowledge and skills needed for successful adaptation to
changing circumstances are continuously acquired over the life cycle. This
framework, and the key competencies needed to function in today’s com-
plex societies, are elaborated in the OECD’s “Definition and Selection of
Key Competences” (DeSeCo) project (OECD 2005b), 1 which identifies
three broad clusters of competencies:
The ability to use tools interactively, including the ability to use language,
symbols, text, knowledge, information, and technology. This requires
Education and the Supply of Skills to the ECA Market 49
more than having access to a tool and the technical skills required to
handle it (e.g., the skills needed to read a text or use software). Rather,
individuals also need to create and adapt knowledge and skills, which
require a familiarity with the tool itself, as well as an understanding of
how it changes the way one can interact with the world and how it can
be used to accomplish broader goals.
The ability to interact in heterogeneous groups, including the ability to
relate well to others, cooperate and work in teams, and manage and
resolve conflicts. As societies become more fragmented and more
diverse, it becomes important to manage interpersonal relationships
well, which benefits both individuals and their abilities to build new
forms of cooperation.
Box 2.1
Defining “Skills,” “Competencies,” and Other Terms

This book utilizes a range of education and learning concepts that were used to
develop the European Qualifications Framework for Lifelong Learning of the EU.
There, certain key terms are defined as follows:
• Skills: the ability to apply knowledge and use know-how to complete tasks and
solve problems. Skills may be cognitive (involving the use of logical, intuitive,
and creative thinking) or practical (involving manual dexterity and the use of
methods, materials, tools, and instruments).
• Competence (or competency): the proven ability to use knowledge; skills; and
personal, social, and methodological abilities in work or study situations and in
professional and personal development. Competence may be described in
terms of responsibility and autonomy.
• Learning outcomes: statements of what a learner knows, understands, and is
able to do on completion of a learning process, defined in terms of knowledge,
skills, and competencies.
• Qualifications: a formal assessment and validation by a competency body that
determines an individual has achieved learning outcomes in conformance with
given standards.
• Knowledge: the outcome of the assimilation of information through learning.
Knowledge is the body of facts, principles, theories, and practices related to a
field of work or study. It can be theoretical or factual or both.
Source: EU 2008.
The ability to act autonomously, including the ability to form and con-
duct life plans and personal projects, as well as to assert rights, interests,
limits, and needs. Acting autonomously does not mean functioning in social
isolation. Instead, individuals must act autonomously in order to partici-
pate effectively in the development of society and function well in differ-
ent spheres of life, including the workplace, family life, and social life.
To develop the higher-level competencies in demand among contem-
porary employers, students must become able to organize and regulate
their own learning, learn both independently and in groups, and over-
come difficulties in the learning process. This versatility requires that they
become aware of their own thinking processes, as well as their learning
strategies and methods. What they acquire from school and university
education is the cognitive and motivational prerequisite for successful
learning in future life, with adult education and training helping refresh
and build on these foundations.
Education and training systems today are assessed against their ability to
develop the above-described competencies in their students. Clearly, such
systems must first provide individuals access to schooling and training, as
measured by years of education of recent graduates, current enrollment
trends, and the share of working-age adults who participate in adult learn-
ing programs. Schooling must, moreover, be of sufficient quality in order for
it to result in actual learning, or the acquisition of knowledge and skills
(sometimes referred to as educational outcomes) by students. Various
national and international student assessment instruments have been devel-
oped precisely to measure these learning outcomes. Finally, an education
system must be able to impart the more complex competencies described
above in graduates at all levels; that is, it must impart education that is
relevant with respect to labor market demands and social conditions.
Formal Education in ECA Countries: High Attainment

and Good Quality Relative to Current Income Levels
Starting from a high base, the ECA region has generally seen a significant
improvement in educational attainment since the start of the transition
toward market economies in the early 1990s. ECA countries went into the
transition with an average of at least eight years of compulsory education;
by 2006, this number had increased to nine years, corresponding roughly
to the primary and lower secondary cycle (see annex 2A, table 2A.1).
This increase closed the gap between most ECA and high-income OECD
countries in terms of average years of education.
The most rapid growth in attainment has taken place at the tertiary
level, albeit with considerable variation across the region. In fact, the
increase in tertiary enrollment rates over the period 1999–2006 is higher
for ECA countries (from below 40 percent to almost 50 percent) than for
either high-income countries or the East Asia or Latin America regions.
This growth has occurred across the region over the past decade, although
Turkey and the South Eastern European and low-income CIS countries
are trailing the EU10+1 and middle-income CIS countries. Gross tertiary
enrollment rates in the ECA region as a whole demonstrate a pattern
similar to those in secondary education: a dip in the mid-1990s, followed
by a strong rebound. For the period 1990–2006 overall, the level of gross
tertiary enrollments in the region were second only to the high-income
countries (see figure 2.1).
ECA countries not only do well in terms of attainment, they also
deliver relatively good-quality education, especially in the early grades.
Indeed, many younger students in the region outperform their Western
European peers in reading at the fourth-grade level. For example, stu-
dents from the majority of participating ECA countries scored above
the scale average2 on the Progress in International Reading Literacy
Study (PIRLS) in both 2001 and 2006. There are, however, important
Figure 2.1 Gross Enrollment Rates in Tertiary Education, by World Bank Region
(percent)
80
70
60
50
40
30
20
10
0
lA d
Af and
cif d
ia
be nd
Af aran
e
ag
ra an
Pa an
As
rib a a
sia
ic
an
er
ric
ric
nt pe
rth ast
ia
h
h
Ca ic
av
Sa
ut
As
Ce uro
e er
No e E
So
b-
CD
st
th m
dl
Su
E
Ea
OE
n
id
tin
r
M
te
La
s
Ea
1990 1999 2006
Source: Edstats Database.

Note: In the case of South Asia, the figure for 1999 is actually for 2000.
variations across the ECA region. Countries that have not joined the EU,
for example, scored below the average, with the marked exception of the
Russian Federation, which scored well above the average and demon-
strated the greatest improvement between 2001 and 2006 (see fig-
ure 2.2). This strong performance in reading at the fourth-grade level
indicates that schools in ECA countries are very capable of providing a
quality education at the early stages.
Many—but not all—ECA countries continue to do well on interna-
tional assessments of reading at or near the end of lower secondary educa-
tion (e.g., Programme for International Student Assessment [PISA],
Trends in International Mathematics and Science Study [TIMSS]) relative
to the performance of other countries at the same income level.3 Figure
2.3 provides a scatter plot of 2008 per capita GDP against the PISA 2009
reading scores of countries participating in PISA (with ECA countries
highlighted by their letter acronyms). The figure shows a positive rela-
tionship between the two values, suggesting that 15-year-olds in countries
with higher national incomes tend to perform better on the reading com-
ponent of the PISA. Significantly, the ECA region seems divided. Certain
countries (mainly the new EU member states but also Turkey and Serbia)
lie above the regression line, indicating that the reading performance of
their students is better than would be predicted on the basis of GDP per
capita alone. Others lie below the regression line, including Romania,
Montenegro, Bulgaria, Azerbaijan, and the Kyrgyz Republic. Students in
these countries scored lower on the reading component of PISA than
their national per capita GDP would have predicted.
Why Are Skills Emerging as a Problem if Education

Given high attainment levels, booming tertiary enrollments, and respect-
able educational quality at lower education levels, why are firms in the
ECA region still complaining? This is a question that cannot be fully
resolved, mainly because crucial information is missing. Unfortunately,
there are no internationally comparable data on the quality of upper
secondary or tertiary education or the relevance of education at these
levels—information that would indicate whether graduates have the right
skills for the modern job market There are international assessments
on student competencies up to the age of 15 (usually the end of lower
secondary education), but few seek jobs at that age. Beyond these early-
stage assessments, no wide-ranging, fully reliable information exists on
Figure 2.2 Assessing a Skill—Reading Performance in the Fourth Grade: PIRLS Performance of ECA Countries, 2001 and 2006
600
550
mean performance
500
450
400
350
300
Macedonia,
FYR
Bulgaria
Czech Republic
Hungary
Latvia
Lithuania
Poland
Romania
Slovak Republic
Slovenia
Georgia
Moldova
Russian
Federation
Turkey
South EU10+1 Low-CIS Mid- Turkey
Eastern CIS
Europe
2001 2006 PIRLS scale average
Source: World Bank staff calculations based on PIRLS test score data downloaded from EdStats Database.
Note: Thirty-five countries from around the world (although with limited participation from Latin American countries and no participation from Sub-Saharan Africa) participated in PIRLS
2001. In 2006, 41 countries participated, 15 of which were Western European countries. (A list of participating countries is available at http://nces.ed.gov/surveys/pirls/countries.asp.)
53
Figure 2.3 Analysis of Reading Competency of 15-Year-Old Students on

the PISA 2009
550
POL EST
500 HUN
LVA SVN
TUR HRV CZE
LTU SVK
450 RUS
SRB
BGR
ROM
MNE
400 KAZ
ALB
AZE
350
KGZ
300
0 20,000 40,000 60,000 80,000 100,000
Sources: PISA 2009 Database and World Bank staff calculations based on World Development Indicators Database.
Note: The figure shows a regression line representing countries’ predicted PISA scores based solely on GDP per
capita, compared to the OECD mean reading score (horizontal line) on the assessment and GDP per capita in
2008. See “Abbreviations” for a key to country abbreviations.
student competencies. The information that is available includes basic

cognitive skills but not behavioral skills, referring to such issues as team-
work and attitudes toward the job, which are increasingly sought after by
employers. For policy makers to better understand the causes of the
emerging skills bottleneck—and how to address it—this information
deficit needs to be closed.
In addition to the data gaps, there are clear signs that ECA’s educa-
tion systems are facing problems related to quality, which are likely to
be part of the reason why firms are complaining about declining skills.
One problem is that too many students in the ECA region are simply not
learning what firms increasingly need. Compared to OECD countries,
for example, ECA countries have a much larger proportion of students
who fail to acquire even the most basic skills. One measure of the mag-
nitude of this problem is the proportion of 15-year-olds who scored at
or below Level 1 on the reading part of PISA 2009 (see figure 2.4). In
the Kyrgyz Republic, 83 percent of students scored at this low level; in
Serbia, the proportion was 32.3 percent.
These students may not be failing according to national guidelines
(and tests) for the competencies that a 15-year-old is expected to acquire,
but their low scores on this assessment indicate serious deficiencies in
their ability to use reading literacy as a tool to acquire knowledge and
skills in other areas—an important competency needed to work in the
Box 2.2
Testing for Reading Competency in the Programme

for International Student Assessment (PISA)
It is important to understand what is meant by competencies in the context of
reading. In PISA, “reading literacy” is defined as understanding, using, and reflecting
on written texts in order to achieve one’s goals, develop one’s knowledge and
potential, and participate in society (OECD 2007a, 284). This definition goes beyond
the traditional notion of decoding information and a literal interpretation of what
is written towards more applied tasks. Reading scores in all PISA cycles are reported
according to five levels of proficiency, corresponding to tasks of varying difficulty,
making it possible not only to rank students’ performance but also to describe
what students can do. Each successive reading level is associated with tasks of
ascending difficulty. Students scoring below 355 points, that is, those who do not
reach Level 1, are not able to routinely show the most basic reading skills that PISA
seeks to measure.
While such performance should not be interpreted to mean that these stu-
dents have no literacy skills, performance below Level 1 does signal serious defi-
ciencies in students’ ability to use reading literacy as a tool for the acquisition of
knowledge and skills in other areas—an important competency. These students
may therefore be at risk not only of difficulties in their initial transition from educa-
tion to work, but may fail to benefit from further education and learning oppor-
tunities throughout life.
global economy. Similar weaknesses can be observed when looking at the

share of 15-year-olds exhibiting serious problems with numeracy skills
(e.g., see box 2.3 on Turkey). Consequently, these students may experi-
ence difficulties if they continue past secondary school and, of perhaps
greater concern, when they attempt to enter the modern workforce. It is
thus likely that when firms complain about workers not having the right
skills, they are, in part, complaining about the fact that too many students
simply did not learn enough during compulsory education.
Another reason why employers perceive a skills shortage may be that
ECA education systems are successful in imparting basic skills, but less
successful in imparting problem-solving skills. For example, it can be
reasonably interpreted from the discrepancy in students’ performance on
the PIRLS and the PISA reading assessments (see figures 2.2 and 2.3,
56
0
10
20
30
40
50
60
70
80
90
Kyrgyz Republic
(percent)
Azerbaijan
Kazakhstan
Albania
Montenegro
Bulgaria
Romania Section of PISA 2009
Russian Federation
Serbia
Turkey
ECA
Lithuania
Czech Republic
Croatia
Slovak Republic
Slovenia
Latvia
Hungary
Poland
Estonia
Mexico
Chile
Austria
Israel
Note: Horizontal line is the average percentage across OECD countries.

Luxembourg
Source: World Bank Staff calculations based on OECD PISA 2009 database.
Greece
Italy
France
Spain
Germany
United Kingdom
Belgium
Portugal
United States
Sweden
OECD, non-ECA
Ireland
Iceland
Switzerland
Denmark
Norway
New Zealand
Netherlands
Australia
Japan
Canada
Finland
Korea, Rep.
Panama
Peru
Qatar
Indonesia
Argentina
Tunisia
Brazil
Jordan
Trinidad and Tobago
Colombia
Thailand
Figure 2.4 Assessing a Competency—Reading Level at Age 15: Share of Students Scoring Level 1 or Below on Reading
Others
Uruguay
Dubai (UAE)
Liechtenstein
Chinese Taipei
Macao SAR, China
Singapore
Hong Kong SAR, China
Shanghai, China
Box 2.3
Turkey: Many Students Fail to Acquire Even the

Most Basic Proficiencies
Learning outcomes are vastly uneven across different types of schools in Turkey,
resulting in lifelong inequities for students, depending on the location of their
schools. While roughly one-third of 15-year-olds in the country had test scores
in mathematics on the PISA 2006 that were above the OECD average of 500,
nearly 50 percent of students achieved only the most basic proficiency (i.e., they
scored below 420). Equivalent disparities were found with respect to reading
and science.
Figure B2.3 Distribution of PISA 2006 Math Scores by Type of School

in Turkey
800 45
700 679 40
600 35
528
500 435 30
percent
536
score
25
400 382 20
387
300 15
200 10
100 5
0 0
s
hi e
ol l
s
ho a
ol
ol
ol
ol
ac siv
sc tion
ho
ho
ho
ho
ng
s
te n
e te
sc
sc
sc
sc
gh ca
ag h in
hi vo
gh
gh
h
ig
ig
hi
hi
nu it
lh
lh
ia
la w
e
na
ra
ol
nc
ia
n ol
ne
at
ol
tio
ig o
ie
re ch
An
at
ge
sc
ca
An
fo h s
vo
g
hi
share of students math mean score
Source: World Bank 2011a.
Well-performing students in Turkey are located in a handful of specialized

(elite) schools across the country (science high schools, high schools with inten-
sive foreign language training, and Anatolian high schools). Access to these types
of schools, which have better teaching and learning conditions, is controlled by
examination and highly correlated with the income and wealth of students’
families. One in two students in science high schools and one in three students
(continued next page)

Box 2.3 (continued)
in the other two types of high schools belong to the richest 20 percent of house-
holds in the country. These asymmetries in access to different types of public
schools, and the variable quality of education that these schools provide, consti-
tute a formidable barrier to improving the quality of education for all in Turkey.
International evidence shows that very selective systems that track students at an
early age, such as the Turkish system, not only significantly increase educational
inequality, but also reduce a country’s average educational performance
(Hanushek and Wößmann 2008).
On a positive note, Turkish students did significantly better on the PISA 2009
test (compared to their performance in 2006) in all tested areas: science, reading,
and math. This improvement may be a first indication that recent reforms, most
notably the 2004 reform in the primary school curricula—which moved to a stu-
dent-centered, constructivist approach—may have started to pay off. Despite the
improvements, though, a significant proportion of 15-year-olds (25 percent in
reading, 30 percent in math, and 42 percent in science) in Turkey continue to
perform below even the most basic proficiency levels. It is critical to remember,
moreover, that the PISA results represent only the 57 percent of 15-year-olds who
are still in the educational system.
Source: PISA 2009 Database; World Bank forthcoming.
respectively) that most ECA countries do well in teaching children to

“learn to read” but less well in teaching them to “read to learn.” These
conclusions are tentative in nature because comparing PIRLS and PISA
test results is not straightforward, in that PIRLS tests fourth-graders,
while PISA assesses 15-year-olds. Nevertheless, one reasonable interpre-
tation from the discrepancy in their performance on these two tests is
that ECA countries do not impart advanced competencies to students as
well as they impart basic reading skills.
As a result, 15-year-olds in the ECA region may be able to read pro-
ficiently, but not adequately enough to access, store, and utilize informa-
tion from that reading, which can be described as functional literacy. An
underlying reason for this underperformance may be that education
systems in the region continue to place too much emphasis on convey-
ing facts and knowledge, instead of teaching problem solving and critical
thinking (see box 2.4 on Russia). Yet it is exactly these latter skills that
more and more firms seem to be seeking.
An additional problem may be that ECA countries are not good at

imparting skills that are in demand but not being tested. For instance,
behavioral skills (discussed in chapter 1) are consistently being demanded
by employers but are apparently not well provided by ECA education
systems. The extent to which they are struggling on this front is difficult
to quantify, however, since tests such as PISA do not measure these skills
(or, for that matter, other skills in high demand, such as knowledge of
information technology or a foreign language).
Box 2.4
Russia: Building Higher-Order Skills Is Proving Difficult

Russian students perform relatively well on the knowledge, or factual, dimension
of international learning assessments. For example, Russia was the top-ranked
country on PIRLS 2006, a test of the reading achievement of grade 4 students.
Russian students were also relatively highly ranked on TIMSS 2007, a curriculum-
based test of math and science achievement of 4th and 8th graders. By way of
contrast, Russian students did relatively poorly on the PISA 2006, which tests a
student’s ability to use knowledge and thinking skills to solve complex problems
in real-world situations. Indeed, detailed analysis of TIMSS results shows that the
scores of Russian students on factual knowledge were significantly higher than
their scores on reasoning using this knowledge.
These results suggest that the traditional strength of the Russian education
system is in conveying facts or knowledge or both, not problem solving or critical
thinking skills. A recent detailed analysis (Tumeneva 2006) shows that the Russian
curriculum still reflects deep traditions of the Soviet curriculum, focusing more on
systematic academic knowledge and less on the application of this knowledge,
or critical reasoning. Russian textbooks, for example, typically use texts that are
preselected to illustrate specific ideas or themes; Russian pedagogy does not
require students to work with real-world situations or original texts. The recent
introduction of a school-leaving exam, which also serves as a centralized univer-
sity entrance examination, has strengthened this orientation. The test places a
heavy emphasis on factual knowledge, which encourages teachers and students
to memorize answers.
These weaknesses in the Russian curriculum and pedagogy were widely
discussed after the results of the PISA 2000 were published, which showed

Box 2.4 (continued)
relatively poor performance by Russian students on higher-order skills. The

Ministry of Education and Science subsequently led two efforts to develop more
“competency-based” standards. Slowly, these efforts have started to show results.
Significant improvements in student scores on TIMSS and PIRLS have been
observed over time, and for the first time Russian students showed improve-
ments in PISA in 2009. Still, the 2009 score (in reading) was below the score
obtained by Russian students in 2000. Even on PIRLS, the biggest improvements
in student scores relate to factual content (Kouznetsova 2009).
Many educators in Russia believe that the school-leaving exam needs to be
changed to emphasize problem solving and critical thinking skills. This would
encourage further changes in standards, teaching, and assessments at the school
level. Changes are also needed at the higher education level, which currently pro-
motes narrow specializations over transferrable skills and key competencies.
Source: World Bank staff assessment, based on ongoing research by the State University Higher School
of Economics, Moscow.
As noted earlier, where it is measured, educational quality in the region

has not consistently improved over the past decade in many ECA coun-
tries. Only a few countries have managed to improve their test results on
all international tests since 1995 (see annex 2A, table 2A.2). The results
of four countries (Bulgaria, the former Yugoslav Republic of Macedonia,
the Czech Republic, and Croatia) suggest that the quality of education
has deteriorated over time, while the majority of ECA countries have
mixed results (some of their test results show improvement over time,
while others show deteriorating results). Three countries—Turkey, Latvia,
and Poland—show steady improvement across all tests and over time.
Moreover, three latecomers to international testing—Serbia, Armenia, and
the Kyrgyz Republic—show improvement over a shorter time horizon
(Serbia and Armenia first participated in international assessments in
2003; the Kyrgyz Republic, in 2006). However, PISA 2009 brought some
positive news: several ECA countries (including Bulgaria, Romania,
Serbia, the Kyrgyz Republic, and Turkey) showed large improvements
from their 2006 results.
Moreover, indirect evidence suggests that the quality of upper second-
ary and tertiary education also may not be improving. Unfortunately,
there is no equivalent of the PISA, TIMSS, or PIRLS for upper secondary
or higher education which would enable objective comparisons of the

quality of student competencies across countries and across time.
Moreover, no ECA country has yet put in place a monitoring system to
track the employment and wage outcomes of recent graduates in order to
assess their employability and the relevance of their education.4 This lack
of data makes it impossible to draw firm conclusions about the quality of
upper secondary and tertiary education. Instead, this book looks at indi-
rect evidence to make inferences.
Limited improvements on international assessments at the end of
lower secondary education (PISA and TIMSS) imply only limited
improvements in the quality of upper secondary education (a finding
that will persist unless upper secondary institutions can make up for the
declining competencies of incoming students). One factor, however, may
have helped improve upper secondary education: fewer students are
enrolled in terminal vocational schools where, in many ECA countries,
the quality of schooling has historically varied a great deal. As the expe-
rience of Poland shows (see box 4.2), providing students with further
general secondary education instead of terminal vocation programs,
which can have a big impact on providing them the types of competen-
cies measured by international assessments.
The fact that the quality of compulsory education has not been reli-
ably improving is discussed further in chapter 3. Here, it suffices to note
that whatever policy measures have been tried in the past approximately
15 years (the period for which international test results are available)
have not worked consistently to improve learning outcomes. Most trou-
bling, the shift to smaller classes and smaller student-teacher ratios
(documented in chapter 3) and the consequent increase in per student
spending do not seem to be delivering better learning outcomes.
At the tertiary level, the indirect evidence is again not encouraging. It
seems probable that the rapid expansion of tertiary education (see fig-
ure 2.1) has resulted in poorer-quality education for a number of reasons.
To begin with, this expansion took place during a period when few coun-
tries had operational quality assurance agencies. When these agencies
finally became operational, several acted swiftly to point out the wide-
spread noncompliance of many new institutions of higher learning with
even minimum quality standards (see chapter 5). Moreover, the new
“mass” nature of tertiary education meant that less qualified students
continued to study. As noted earlier, when as many as a third of students
at the age of 15 are functionally illiterate in Serbia (according to the 2009
PISA results), such students will unquestionably not graduate with the
same level of proficiency from tertiary education as did the smaller, more
qualified university cohorts of the past.
In addition, a great deal of the expansion of tertiary education has
occurred in new institutions whose short-term motives may be driven more
by profits than a need to guard their institutional reputations. Although
private, profit-based tertiary education providers can create a more flexible,
responsive university system (see chapter 5), there are signs of persistent
problems in this sector, including mounting quantitative (as well as anec-
dotal) evidence of fraud, corruption, and unethical behavior—the scope of
which is extensive—raising doubts about its capacity to ensure that stu-
dents graduate with the requisite skills (and not just diplomas). For example,
a recent survey of university students in several ECA countries found that
more than 60 percent of respondents knew other students who had pur-
chased either entrance to the university or a specific grade (see figure 2.5).
Finally, a large part of the expansion of tertiary education took place
in part-time (or weekend) and long-distance learning programs. The
format of these programs was new and their quality unknown. In 2009,
for example, half of Poland’s undergraduate students were enrolled in
weekend programs. In Romania in the same year, 42 percent of students
were enrolled in private universities, 66 percent of whom were part-time
or long-distance students. It is impossible to determine whether these
Figure 2.5 Students Aware of Bribery for Grade or Exam in their Faculty
90
80
70
60
percent
50
40
30
20
10
0
Bulgaria Croatia Moldova Serbia Kyrgyz
Republica
Source: Heyneman, Anderson, and Nuraliyeva 2008.

Note: Data are from a random sample of university students.
a. Data for the Kyrgyz Republic are based on a slightly different question: students were asked if they had “per-
sonal experiences with corruption and bribery in their university.” The value reported in the figure is the median
value of Kyrgyz-owned universities reported in table 3 of Heyneman, Anderson, and Nuraliyeva 2008.
students are receiving the same quality education as their peers in full-
time, regular programs.
Students May Not Be Acquiring the Right Skills

It is not only probable that the quality of graduates’ problem-solving and
behavioral skills are inadequate, but students at both the secondary and
tertiary levels may also be graduating with the wrong set of skills. At the
secondary level, for example, the collapse of vocational schools may have
been too abrupt, stripping students of a viable option to obtain technical
skills that are in strong demand. At the tertiary level, too many students
may be enrolling in low-cost programs (that is, the only programs avail-
able to them), rather than programs that would meet current and future
employer needs.
Vocational school systems in the ECA region rapidly declined in the
early years of transition for a number of reasons, as students fled from
these schools in large numbers (see box 2.5 for a brief history of the
vocational school system in the region.) This decline happened in virtu-
ally all ECA countries (see figure 2.6), with the noticeable exceptions of
Uzbekistan, Serbia, Croatia, the Czech Republic, Moldova, and Turkey. In
some countries—Belarus and Armenia, for instance—vocational school
systems have become virtually nonexistent.
Students abandoned vocational schools because these schools failed
to adjust to new realities. Nevertheless, there is evidence to suggest that
firms still seem to want vocational and technical skills. For instance, as
mentioned in chapter 1, students who finish school with vocational or
technical skills—especially in middle-income CIS countries—are less
likely to end up unemployed than students with only a general second-
ary education (Turkey being an exception). This finding has been cor-
roborated by recent surveys of graduates regarding the school-to-work
transition in Serbia and Ukraine (the two only countries where such
surveys have been conducted). As table 2.1 shows, it seems easier for
students who have technical or vocational training to transition into a
job than for those who have only a general education. In Serbia, one-
third of graduates of a three-year vocational training course find their
first job within six months, compared with only one-fourth of gradu-
ates of general secondary schools. In Ukraine, less than 30 percent of
graduates of secondary technical schools look for their first job for
more than a year, compared to 44 percent of graduates of general sec-
ondary schools.5
Box 2.5
The History of Vocational Education in the ECA Region

Historically, secondary education has been heavily weighted in favor of voca-
tional training in the ECA region. The former planned economies entered the
transition with a very high percentage of upper secondary school students
enrolled in vocational programs—averaging 61.3 percent of all such students in
1989 (see figure B2.5 below). Most of these programs were terminal, that is, they
did not provide students access to higher education. Most also involved applied
training in nearby public enterprises, which typically recruited many program
graduates. The most highly vocationalized programs were in Central and Eastern
Europe, where the average share of vocational secondary school students was
over 70 percent. In Turkey, the single nontransition country in the ECA region,
secondary education also has a strong vocational focus, with programs patterned
after the German dual-system model.
Figure B2.5 Global Comparison of Average Vocational Enrollment of

Upper Secondary Students by Region, 1989, 1999, and 2007
(percent)
percentage of upper secondary students
70
enrolled in vocational programs
60
50
40
30
20
10
0
A
Af and
cif d
tin Asia d
er tral
Af aran
pe
Pa an
t n
EC
ro
es a
a
ic
ica
ica
Am en
ric
rth ast
ia
W th
Eu
r
Sa
C
As
u
No E
rn
d
So
b-
e
st
an
te
dl
Su
Ea
es
id
W
M
La
1989 1999 2007
Source: UIS (UNESCO Institute for Statistics).

Note: Percentages represent averages for those countries for which data was available.

Box 2.5 (continued)
The inherited vocational focus of secondary education in ECA countries

reflects the incentives of planned economies and the strong linkages that such
economies forged between training institutions and state-owned enterprises.
Since general secondary education was seen primarily as a preparation for higher
education, enrollments in this track were limited to the few students who were
expected to progress to university.
Vocational education in the previous system was thus the primary means of
training workers for specific jobs in public enterprises. These enterprises typically
contributed to the curriculum design of vocational education programs, provided
instructors and facilities for practical training, and recruited many graduates, often
for lifetime jobs. Labor hoarding proved an effective approach for ensuring that
enterprises could meet planned production quotas.
Economic transition has brought new requirements that necessitate wide-
spread changes in vocational education programs, but shrinking budgets have
made it hard to carry out these changes, particularly the hiring of instructors and
curriculum designers with needed skills. The move to a competitive market econ-
omy has also profoundly changed the setting in which firms operate—hard bud-
get constraints and falling sales have put an end to labor hoarding and created
widespread unemployment. These conditions have made it impossible for most
enterprises to continue to offer training to vocational school students.
Source: Authors.
In sum, the quality and relevance of vocational education in the ECA

region appear to be deteriorating at the same time that firms are increas-
ing their demands for higher-level technical skills.
Adult Learning Is Limited in the Region

Demographic trends in the ECA region imply that young people entering
the job market directly from the formal schooling sector make up a
smaller proportion of the workforce today than just 10 or 20 years ago, a
trend that will continue. When firms complain of a skills shortage, they are
not referring solely to the quality of new labor market entrants; they are
also indicating that the skills they need cannot be found among all workers
in the economy. Given the aging of the population in many countries, it
66
0
10
20
30
40
50
60
70
80
90
100
(percent)
Source: UIS.
Hungary
Lithuania
Estonia
Latvia
Poland
Note: LI = low-income, MI = middle-income.

Bulgaria
EU10+1
Slovenia
Romania
Croatia
Slovak Republic
Czech Republic
1999
Albania
Macedonia, FYR
SEE
1989 Serbia
Belarus
Kazakhstan
2007
MI CIS
Ukraine
Russian Federation
Azerbaijan
Armenia
Figure 2.6 Share of Upper Secondary Students Enrolled in Vocational Programs in ECA Countries over Time
Georgia
Tajikistan
LI CIS
Kyrgyz Republic
Moldova
Uzbekistan
Turkey
Turkey
Table 2.1 Time Needed to Find First Job by Level of Educational Attainment
in Serbia and Ukraine
(percent)
a. Serbia
Secondary Secondary Secondary
Primary trade vocational general College University
No search 7.0 5.7 4.7 9.6 7.7 11.7
1–3 months 22.0 25.7 22.3 19.2 23.1 26.0
4–6 months 8.4 7.5 7.2 6.7 9.1 12.8
7–12 months 10.6 10.8 12.4 8.7 18.0 9.4
1–2 yrs 15.0 15.2 19.5 8.4 13.2 21.6
> 2 yrs 37.0 35.1 33.9 47.3 29.0 18.4
Total 100 100 100 100 100 100
Source: Huitfeldt, Johansen, and Kogan 2008.
b. Ukraine
Less than Secondary Secondary Junior
secondary vocational general specialist University
No search 0.9 7.3 4.8 9.0 26.6
1–3 months 32.7 49.5 31.4 58.2 44.4
4–6 months 6.7 7.6 10.5 7.4 8.4
7–12 months 7.5 6.9 9.6 6.1 5.2
1–2 yrs 5.2 11.8 9.1 4.5 5.4
> 2 yrs 47.0 16.9 34.7 14.7 10.0
Total 100 100 100 100 100
Source: Huitfeldt, Johansen, and Kogan 2008.
will become increasingly important to ensure that all workers—not just

those recently out of school—continue to upgrade their skills.
As much as one-half of the skills and knowledge that an individual
accumulates over his or her lifetime is gained during the period after
formal schooling (Heckman 1999). A large international literature docu-
mented by empirical evidence—largely for OECD countries—shows that
adult education and training advances labor market outcomes, including
higher wages and employment. This type of training also has a strong
positive effect on productivity and a high internal rate of return (see, for
example, OECD 2004a and Dearden, Reed, and van Reenen 2000).
Significant evidence from countries around the world also documents
that market failures tend to lead to underinvestment in adult education
and training on the part of both individuals and firms; here, the countries
of ECA are no exception (see chapter 6).
In fact, many countries in the ECA region have not yet begun to ade-
quately promote adult learning as a means of addressing the current skills
deficit and improving worker productivity in the face of declining demo-
graphics. Although a substantial share of workers participate in adult
education and training programs in several ECA countries, the percent-
ages vary across the region. Even in countries that have relatively high
availability of continuing vocational education and training programs, the
people who take part in these programs tend to be already skilled workers,
rather than nonskilled and “nonproductive” workers. Finally, it seems that
in many ECA countries, programs to help retrain or reskill the unem-
ployed only reach a few of the unemployed.
Adult learning programs (see box 2.6) have remained a blind spot in
education and training policy across much of the ECA region. A key
component of the socialist production process, worker training has under-
gone substantial changes as a result of the transition to market economies.
Under communism, such training focused exclusively on continuing
vocational training of the employed workforce. With the exception of
socialist Yugoslavia, unemployment in these countries was nonexistent by
definition, hence adult education and training of the unemployed was not
a feature of their education and training systems.
Partly, adult education and training remains a blind spot in the region
because remarkably few data on the sector are available. While several
OECD countries have been conducting national surveys of adult train-
ing in firms for many years, such surveys have not been comparative
across countries and data for ECA countries is sparse. Even OECD
countries focus on the quantity of this learning (i.e., number of partici-
pants, hours of learning), not its quality and relevance. For the ECA
region as a whole, the 2005 and 2008 Business Environment and
Enterprise Performance Surveys (BEEPS 2005, 2008) included some
limited questions on adult training by firm size and some degree of
worker characteristics; however, problems relating to cross-country
comparability make it difficult to reach solid conclusions. In the EU, the
firm-based Continuing Vocational Training Surveys (CVTS) of 1999
and 2005, which offer information on worker training, and the Adult
Education Survey (AES) of 2007, which offers data on individual par-
ticipation in formal or nonformal education and training, are the first
surveys that permitted cross-country analysis of adult education and
training patterns. Both included a number of ECA countries and serve
as the basis of much of the analysis in this book. The data used here,
therefore, center disproportionately on the EU10+1.
Box 2.6
Two Forms of Adult Learning

Two categories of adult education and training are addressed in this report; both
categories aim at adults between 25 and 64 years of age (for further discussion,
including on the financing of adult education and training, see chapter 6):
Continuing Vocational Education and Training (CVET). These programs are aimed
at people who are currently employed. First, they offer in-service training designed
to aid employees in acquiring new competencies or improving existing ones rel-
evant to their firms’ operations. In-service training plays a critical role in increasing
the human capital of the existing workforce, addressing skills depletion, and keep-
ing older workers productive for longer, as well as alleviating skills mismatches.
Second, these programs offer education and training to individuals who seek to
develop skills that will raise their chances of moving to a better job (that is, educa-
tion and training not related to their current job).
Retraining and remedial basic skills training (“second-chance” education). This

type of training aims at the nonemployed, that is, the unemployed and people
outside the labor force who face skills-related barriers to employment. In many
ECA countries, high percentages of the population continue to lack basic skills as
a result of leaving school early or poor learning achievement. Most ECA countries
provide training as part of active labor market programs and there is reason to
believe that well-designed programs can have positive economic returns
(Betcherman, Olivas, and Dar 2004). In addition, “second-chance” education pro-
grams address (basic) skills shortages, including literacy, in order to help youths
and adults access the labor market and further their education and training.
ECA Lags Behind West European Countries in

Continuing Education and Training
Today, the provision of continuing vocational education and training
(CVET) in firms varies considerably across the ECA region. Figure 2.7
presents data from the 2008 BEEPS survey in ECA countries, showing
the share of manufacturing firms that offered formal training programs to
full-time employees. The variation is wide between the Czech Republic
and Estonia, where about 70 percent of firms provide such training, and
Azerbaijan, where only about 10 percent do. With few exceptions, the
70
Figure 2.7 Share of ECA Firms that Offer Formal Training Programs for Permanent, Full-Time Employees, 2008
80
69 71
70 66
61
60
52
50 46 47 44
43
percent
41
40 37
33 33
31 30 30
30 28 29
25 25 25 25
21
19 20
20
15 15
10 11
10
0
HUN
ROU
BGR
LVA
LTU
SVN
POL
EST
CZE
MKD
ALB
Kosovo
MNE
TUR
SRB
BHR
UKR
KAZ
BLR
RUS
UZB
AZE
GEO
KGZ
ARM
MDA
HRV
SVK
TJK
EU10+1 SEE MI CIS LI CIS
Source: World Bank Staff calculations based on raw dataset for EBRD-World Bank 2009.
Note: LI = low-income, MI = middle-income. See “Acronymvs and Abbreviations” for a key to country abbreviations.
distribution broadly follows a GDP per capita pattern, with low-income

CIS countries on one side and advanced EU10+1 countries on the other.
Independent of the data source, the Czech Republic, Slovenia, Estonia,
and the Slovak Republic appear to lead the ECA region in adult education
and training. Many low-income CIS countries occupy the bottom end of
the distribution, with the bulk of South Eastern European and middle-
income CIS countries in between. However, there are some surprises at
both the top and bottom end of the distribution. Bosnia and Herzegovina
is in the top half, among EU10+1 countries, while Hungary, Romania, and
Bulgaria—despite EU membership—are in the bottom half.
Data from the 2005 CVTS show that more workers in the Czech
Republic and Slovenia participate in training proportionately than else-
where in the EU (see figure 2.8, panel a). Interestingly, results in countries
that are thought to be similar, such as the Baltic states, can differ signifi-
cantly. For example, Estonia shows consistently higher training participa-
tion than Lithuania and Latvia. Individual-level data from the 2007 AES
confirm this picture (figure 2.8, panel b).
However, several country studies have shown that while the share of
firms that offer training may be high in many ECA countries, the share of
actual workers who participate in training is relatively low and the hours
of training are limited. A recent enterprise survey in Russia, for example,
revealed that while 58 percent of firms in a pooled sample of small,
medium, and large firms conducted training, only 7.7 percent of skilled
and 1.4 percent of unskilled workers actually participated in the training
(Tan et al. 2007).6
Figure 2.9 (panel a) presents the percentage of EU workers participat-
ing in (rather than firms offering) continuing vocational training in those
firms that offer training. In the Czech Republic, close to 70 percent of
workers in firms that offer continuing vocational training participate in
training activities, whereas the equivalent share in Hungary is less than 25
percent. Moreover, panel b shows that with the exception of the Czech
Republic, the Slovak Republic, and Slovenia, workers in the EU10 coun-
tries of the ECA region participate in such training for fewer hours than
their EU peers. This implies that the productivity effects of worker train-
ing may be lower in ECA countries than in Western European countries
(although obviously the productivity effect also depends on the quality
and relevance of the training, which is not captured here).
Compared to other countries around the world, the ECA region fares
similarly in terms of uneven participation in CVET, which is concen-
trated among workers with the best labor market opportunities. According
Figure 2.8 Employee- and Firm-level Data on Worker Participation in CVET

Courses, EU10 and other EU Member States
(percent)
a. Percentage of employees of all enterprises who participate in CVET, 2005

60
50
40
percent
30
20
10
0
EU27
CZE
SVN
EST
POL
ROU
HUN
BGR
LVA
LTU
LUX
FRA
SWE
BEL
FIN
NLD
ESP
AUT
GBR
MTL
DEU
CYP
ITA
NOR
PRT
SVK
DNK
EU10 other EU
b. Percentage of individuals in formal or nonformal

80 education and training, 2007
70
60
50
percent
40
30
20
10
0
HUN
POL
LVA
LTU
BGR
EST
GRC
ITA
ESP
FRA
CYP
AUT
GBR
DEU
NOR
FIN
SWE
SVK
EU10 other EU
Source: World Bank staff calculations based on CVTS 2005 (a) and AES 2007 (b).
Note: See “Abbreviations” for a key to country abbreviations and groupings.
to BEEPS 2005 data, skilled workers in all ECA countries are more likely
to participate in education and training than other workers in the same
firm. Figure 2.10 shows that overall, firm-based worker training data
mask substantial differences in training rates between skilled labor on
one hand and nonskilled and so-called nonproductive workers on the
other, with the latter two groups having a lower (sometimes much
lower) participation rate than that of skilled workers. This finding is
Figure 2.9 Training Length and Percentage of Employees Participating in CVET

Courses, EU10 and EU27
(percent)
a. Percentage of employees participating in CVET courses

80 in enterprises that offer training, 2005
70
60
50
percent
40
30
20
10
0
HUN LVA LTU ROU EST BGR POL EU27 SVK SVN CZE
b. Hours in CVET courses per

16 employee (all enterprises), 2005
14
12
10
hours
8
6
4
2
0
CZE SVN SVK EU27 EST HUN POL LTU ROU BGR LVA
Source: World Bank staff calculations based on CVTS 2005.

Note: See “Abbreviations” for a key to country abbreviations.
consistent with the patterns of adult education and training around

the world.
The extent of worker training also varies between large and small
firms, with the variation generally larger in ECA countries than in EU
countries that have highly developed economies. Large firms worldwide
are substantially more likely to invest in worker training than small firms.
This tendency may provide one explanation for the relatively lower adult
training rates in many ECA countries compared to EU countries with
more advanced economies: in the former, the enterprise sector consists
primarily of small firms—the very kind that are less likely to make train-
ing available.
74
Figure 2.10 ECA Firms that Offer Formal Training to Employees, by Employee Category, 2005
(percent)
90
80
70
60
percent of firms
50
40
30
20
10
a
B
R
Z
U
O
M
DA
S
R
N
KD
Z
A
T
N
V
K
K
B
PO
BE
AZ
RU
CZ
ES
TU
BG
UK
UZ
AL
KA
TJ
KG
SV
HR
LV
RO
LT
BI
HU
SV
GE
AR
SR
M
M
skilled workers unskilled workers nonproductive workers
Source: World Bank staff calculations based on EBRD-World Bank 2005.

Note: See “Abbreviations” for a key to country abbreviations.
a. Results for Serbia (SRB) include Montenegro.
Figure 2.11 (panel a) presents 2005 BEEPS data that confirm the
internationally well-established fact that large firms provide more training
than small firms. The bottom panel presents data from the 2005 CVTS
on the best- and worst-performing ECA countries on this measure—the
Czech Republic and Bulgaria, respectively—and compares this data with
the best-performing EU country in the CVTS sample, the United
Kingdom. In Bulgaria and the Czech Republic, there is a wide gap in the
share of companies that offer training between small and large enter-
prises, although at different levels, while the United Kingdom shows little
divergence between small and large firms.
Very Limited Training Programs for the Unemployed

Although retraining, if well designed, can be important for promoting
employment, the actual use of this intervention for the unemployed in
many parts of ECA appears limited. The introduction of passive and
active employment policies in ECA at the start of the transition gave
priority to interventions that retrained and prequalified workers who had
lost their jobs as part of privatization and enterprise restructuring. Today,
retraining programs are a well-established component of active labor
market policies (ALMPs) throughout the ECA region. While there is no
comparative data across the entire region, table 2.2 shows that training
participants represent sizable shares of overall ALMP participants in the
EU10 countries, in some cases, over 50 percent (similar to the EU27 aver-
age). However, not all unemployed participate in ALMPs, and training
actually reaches few unemployed overall in EU10 countries compared to
the EU27 average, with the exception of Slovenia.
Moreover, it appears that “second-chance” education programs form
a relatively small part of the adult education and training system in
many ECA countries. These programs are a key tool for helping early
dropouts and workers who have lost their jobs in mid-career improve
their skills and re-enter the labor market. There is currently no ECA-
wide, individual-level data that would allow for a breakdown of par-
ticipation in second-chance programs by worker characteristics.
However, 2007 AES data show that EU10 countries appear to have
fewer individuals with lower schooling levels in adult education and
training programs than do most older member states (see figure 2.12,
panel a). The EU10 also appear to have substantially fewer inactive
workers and unemployed workers participating in nonformal education
and training programs than most of their EU neighbors (see figure 2.12,
panel b).
Figure 2.11 Large and Small Firms that Offer Training to Skilled Employees,
ECA, EU, and United Kingdom, 2005
(percent)
a. Large vs. small firms: Share of firms that offer

100 training to skilled employees, by firm size
90
80
70
percentage of firms
60
50
40
30
20
10
0
Ba
O
B
TU E
R
RUR
AR S
M
U
Z
M U
M A
HUD
N
SR R
Z
A
H
L
T
B
L
E
N
SV V
K
K
PO
BE
AZ
CZ
ES
BG
UK
UZ
AL
TJ
KA
KG
SV
HR
D
LV
RO
LT
BI
K
GE
small firms medium firms large firms
b. Training enterprises as % of all

enterprises by size, 2005
100
90
percentage of all enterprises
80
70
60
50
40
30
20
10
0
Bulgaria Czech Republic United Kingdom
total 10 to 49 50 to 249 250 or more
Sources: (a) World Bank staff calculations based on raw dataset of EBRD-World Bank 2005. (b) World Bank staff
calculations based on CVTS 2005.
Notes: BEEPS defines small firms as fewer than 50 employees, medium firms as between 50 and 249 employees,
and large firms as more than 250 employees. See “Abbreviations” for a key to country abbreviations.
a. Results for Serbia (SRB) include Montenegro.
Table 2.2 Training Participants as Share of Participants in Active Labor Market

Policies and of Total Unemployed, EU27 and EU10, 2006
Training participants as Training participants as

% of ALMP participants % of total unemployed
EU27 33.6 19.8
Bulgaria 10.9 3.8
Czech Republic 12.6 2.0
Estonia 58.4 2.8
Latvia 52.3 6.3
Lithuania 41.0 7.9
Hungary 19.9 4.4
Poland 21.8 4.2
Romania 18.0 2.1
Slovenia 56.8 19.7
Slovak Republic 1.6 0.6
Source: World Bank staff calculations based on LFS database (Eurostat).
Indeed, anecdotal evidence suggests that remedial basic education

and training represent a small share of the retraining programs of pub-
lic employment services in these countries. (Such programs are often
coupled with nongovernmental organization [NGO]-provided social
inclusion activities). However, several new EU member states have
launched second-chance education and training programs, including
components on literacy and functional literacy, as part of programs
funded by the European Social Fund.
Summary
This chapter has sought to understand why employers in the ECA region
increasingly complain of a skills shortage despite high levels of educa-
tional attainment, respectable educational quality, and booming tertiary
enrollments in ECA countries. Although the arguments presented here
can necessarily be based only on the data currently available, the findings
nevertheless point to problems with the limited quality and relevance of
education as contributing sources of the skills deficit. Graduates of upper
secondary and tertiary education appear to either not be learning enough
while in school or acquiring skills that are not in demand on the labor
market. In addition, the decline of vocational schools at the upper sec-
ondary level may have intensified the skills shortage.
Figure 2.12 Adult Education and Training Program Participants in EU, 2007
a. Participants by education level

100
90
80
70
60
percent
50
40
30
20
10
0
HUN
POL
LTU
LVA
BGR
EST
GRC
ITA
CYP
ESP
FRA
AUT
DEU
GBR
FIN
NOR
SWE
SVK
EU10 other EU
pre-primary, primary, and lower secondary education—ISCED 0-2

upper secondary and postsecondary nontertiary education—ISCED 3-4
tertiary education—ISCED 5-6
b. Participants by labor market status

90
80
70
60
percent
50
40
30
20
10
0
HUN
POL
LVA
LTU
EST
BGR
GRC
ITA
ESP
FRA
CYP
AUT
DEU
GBR
NOR
FIN
SWE
SVK
EU10 other EU
employed unemployed inactive
Source: Bank staff calculations based on AES 2007.

Note: ISCED = International Standard Classification of Education. See “Abbreviations” for a key to country
abbreviations.
International learning assessments show that education systems in the

ECA region are successful in imparting basic skills, but less successful in
imparting the higher-order skills sought by employers. These systems
continue to place too much emphasis on conveying facts and knowledge,
instead of problem solving and critical thinking. The respectable average
performance of ECA countries on international assessments, moreover,
masks a long tail of underperformers: 15-year-olds who have such poor
numeracy and literacy skills that their success in the modern workplace
is highly doubtful.
The quality of education in ECA countries, particularly at the lower
levels, is good compared to countries at the same income level, but unfor-
tunately, this quality is not improving. In fact, it appears to be worsening
at the lower secondary level in many countries, as seen in the stagnant or
deteriorating scores of ECA countries on international assessments at this
stage. At the upper secondary and tertiary levels, educational quality is
less known. Because the tertiary sector expanded significantly without (in
many—though not all—countries) strong quality assurance programs and
information that parents and students need to make informed choices,
the quality of many university degrees is suspect. Qualitative evidence,
moreover, highlights significant unethical behavior in university admit-
tance, grading, and graduation policies.
Underdeveloped adult learning systems in the region threatens to
cement the skills deficit in the region. Given that the populations of
many ECA countries are declining, the need to retrain and reskill adult
workers—as well as to train unemployed workers—is becoming impera-
tive. Yet many countries in the region have only begun to lay the ground-
work for modern systems of adult learning.
In sum, education systems in the region impart excellent founda-
tional skills, but falter at the secondary level, when students need to
acquire the more advanced competencies needed to work in modern,
globalized economies. Despite rapid expansion, the tertiary sector does
not yet appear to be delivering the skills sought by employers, and adults
have very limited educational opportunities to retrain or improve exist-
ing skills.
The next chapter takes a detailed look at the sources of the underper-
formance of ECA education systems and offers a framework for how to
reform these systems in order to enhance educational quality and rele-
vance. Subsequent chapters then explore recommendations for improv-
ing specific levels of the education system, including continuing adult
education and training.
Annex 2A: Education Systems in ECA Today

Table 2A.1 Duration of Education in ECA Countries, Various Years
1991 2000 2006
Albania Primary school starting age 6 6 6
Primary duration 8 4 4
Lower secondary duration .. 4 4
Upper secondary duration .. 4 4
Compulsory years — 8 8
Age after compulsory education 14 14
Armenia Primary school starting age 7 7 7
Lower secondary duration 5 5 5
Upper secondary duration 2 2 2
Age after compulsory education — 15 15
Azerbaijan Primary school starting age 7 6 6
Belarus Primary school starting age 6 6 6
Bosnia and Herzegovina Primary school starting age 7 6 6
Compulsory years — .. ..
Age after compulsory education — — —
Bulgaria Primary school starting age 6 7 7
Croatia Primary school starting age 7 7 7

Table 2A.1 (continued)

1991 2000 2006
Czech Republic Primary school starting age 6 6 6
Estonia Primary school starting age 7 7 7
Georgia Primary school starting age 6 6 6
Hungary Primary school starting age 6 7 7
Kazakhstan Primary school starting age 7 7 7
Kyrgyz Republic Primary school starting age 7 7 7


1991 2000 2006
Latvia Primary school starting age 7 7 7
Lithuania Primary school starting age 7 7 7
Macedonia, FYR Primary school starting age 7 7 7
Moldova Primary school starting age 7 7 7
Montenegro Primary school starting age — — —
Primary duration — — —
Lower secondary duration — — —
Upper secondary duration — — —
Compulsory years — — —
Poland Primary school starting age 7 7 7
Lower secondary duration — 2 3
Upper secondary duration — 4 3
Romania Primary school starting age 6 7 7
Russian Federation Primary school starting age 7 7 7

1991 2000 2006
Serbia Primary school starting age — — —
Primary duration — — —
Lower secondary duration — — —
Slovak Republic Primary school starting age 6 6 6
Slovenia Primary school starting age 7 7 6
Tajikistan Primary school starting age 7 7 7
Lower secondary duration — 5 5
Turkey Primary school starting age 6 6 6
Turkmenistan Primary school starting age 7 7 7
Ukraine Primary school starting age 6 7 6


1991 2000 2006
Uzbekistan Primary school starting age 6 7 7
Source: EdStats Database.
Note: .. = negligible; — = not available. For all countries, compulsory education includes primary and lower sec-
ondary education, with the exception of the following: Lithuania, where compulsory education includes only
part of lower secondary; Azerbaijan, Kazakhstan, and the Russian Federation, where compulsory education also
includes upper secondary; Belarus, the Czech Republic, Hungary, the Slovak Republic, Slovenia, Turkey, and
Turkmenistan, where compulsory education also includes part of upper secondary in at least some of the years;
and Ukraine and Uzbekistan, where the number of compulsory years extends beyond upper secondary
education.
Table 2A.2 ECA Country Results on International Assessments since 1995
TIMSS PISA PIRLS
Math grade 8 Math grade 4 Science grade 8 Science Math Science Reading Reading *
Country
Name 1995 1999 2003 2007 2003 2007 1995 1999 2003 2007 2003 2007 2000 2003 2006 2009 2000 2003 2006 2009 2000 2003 2006 2009 2001 2006
Bulgaria 511 476 464 518 479 470 430 413 428 448 434 439 430 401 429 550 547 Down
Macedonia,
FYR 447 435 458 449 381 401 373 442 442 Down
Czech
Republic 564 520 504 486 574 539 539 515 498 516 509 493 511 523 512 500 492 489 482 478 537 Down
Croatia 467 460 493 486 477 476 Down
Romania 482 472 475 461 486 472 470 462 414 427 418 428 395 424 512 489 Mixed
Slovak
Republic 547 534 508 496 544 535 517 526 498 492 497 495 488 490 469 466 477 518 531 Mixed
Hungary 537 532 529 517 529 510 554 552 543 539 530 536 488 490 490 490 496 503 503 503 480 482 482 494 543 551 Mixed
Lithuania 477 482 502 506 534 530 476 488 519 519 512 514 486 477 487 491 470 468 543 537 Mixed
Slovenia 541 530 493 501 479 502 560 533 520 538 490 518 504 501 518 512 494 483 502 522 Mixed
Moldova 469 460 504 459 472 496 492 500 Mixed
Russian
Federation 535 526 508 512 552 544 538 529 514 530 526 546 478 468 475 468 460 489 479 478 462 442 439 459 528 565 Mixed
Albania 381 377 376 391 349 385 Mixed
Azerbaijan 475 431 382 373 352 362 Mixed
Estonia 531 552 514 512 531 528 500 501 Mixed
Montenegro 399 403 411 401 391 408 Mixed
Serbia 477 486 468 470 437 435 442 436 435 443 412 401 442 Up
Turkey 429 432 433 454 423 423 445 434 423 454 441 447 464 449 Up
Latvia 505 508 536 537 485 503 512 532 542 463 483 486 482 460 489 489 494 458 491 479 484 545 Up
Poland 470 490 495 495 483 498 497 508 479 497 507 500 519 Up
Armenia 478 499 456 500 461 488 437 484 Up
Kyrgyz
85
Republic 310 331 322 330 284 314 Up

86
TIMSS PISA PIRLS
Math grade 8 Math grade 4 Science grade 8 Science Math Science Reading Reading *
Country
Name 1995 1999 2003 2007 2003 2007 1995 1999 2003 2007 2003 2007 2000 2003 2006 2009 2000 2003 2006 2009 2000 2003 2006 2009 2001 2006
Bosnia and
Herzegovina 456 466
Kazakhstan 549 533 405 400 390
Ukraine 462 469 485 474
Georgia 410 438 421 418 471
Tajikistan
Turkmenistan
Uzbekistan
Belarus
Kosovo
Sources: IEA 1995, 1999, 2003, 2007; PISA Databases 2000, 2003, 2006, 2009; IES 2001, 2006.
Note: * Authors’ summary assessment of trends in countries’ scores across time and across various tests.
Notes
1. Along lines similar to the OECD DeSeCo project, the European Qualifications
Framework for Lifelong Learning identifies eight key competences: (1) com-
munication in the mother tongue; (2) communication in foreign languages;
(3) mathematical competence and basic competence in science and technology;
(4) digital competence; (5) learning to learn; (6) interpersonal, intercultural,
and social competences, together with civic competence (7) entrepreneurship;
and (8) cultural expression. In addition, critical thinking, creativity, initiative
taking, problem solving, risk assessment, decision making, and managing
feelings constructively are seen as playing a role in all eight key competencies
(EU 2008). The EU framework identifies the essential knowledge, skills, and
attitudes related to each.
2. This is a scale metric established by setting the mean scores of participating
countries to 500 and the standard deviation to 100, thus enabling compari-
sons over time, since all cycles are placed on this metric so that scores are
equivalent from cycle to cycle. In contrast, the international average, obtained
by averaging across the mean scores for each of the participating countries,
needs to be recomputed for each new cycle, based on the set of participating
countries and changes from cycle to cycle, depending on the set of countries
taking part.
3. Income level is used as a rough measure of the quality of all nonschool factors,
including parental inputs, that help students learn.
4. Some ECA countries are making rapid progress in addressing this information
gap (see chapter 5).
5. As noted in chapter 1, footnote 6, care should be taken when comparing
employment outcomes of vocational vs. general secondary graduates. In par-
ticular, it can be argued that better employment outcomes of workers with
technical secondary education compared with those of workers with general
secondary education reflect the different composition of both groups. One
hypothesis is that due to the selection process, graduates of general secondary
schools who do not proceed to tertiary education are less able, on average,
than the graduates of secondary technical schools.
6. The higher incidence of training among skilled than among nonskilled work-
ers is consistent with evidence from across ECA and the OECD.
CHAPTER 3
Resolving the Skills Shortage in the

ECA Region: A Policy Framework
This chapter identifies the major impediments that prevent the educa-
tion system from delivering the skills that are being demanded by the
labor markets as identified in the previous chapters. It then offers a policy
framework and policy directions to help overcome these impediments.
The impediments and a policy framework for addressing them guide the
discussion of education system reform for the remainder of the book.
Here we argue that the three most pressing problems of ECA educa-
tion systems that get in the way of imparting the needed skills are:
(1) the lack of data on the skills and competencies that students actually
acquire as a result of the educational process; (2) the legacy of central
planning—particularly its effect on the management of education sys-
tems in the region—which makes education systems inflexible and resis-
tant to the initiative of front-line actors (who can actually improve
education); and (3) the inefficient use of resources, a problem that has
become particularly acute in the current economic climate. Certainly
these are not the only impediments to making education systems more
responsive to skills needs in the region. However, they are common to all
ECA countries and lie at the heart of addressing the skills deficiencies
discussed in chapter 2.1
89
The lack of data on student performance, the legacy of central

planning, and the inefficient use of resources affect all levels of education
in the ECA region, but are most apparent at the pre-university level. This
level is the foundation of the education system in every country and gen-
erally absorbs two-thirds of total education funding in the ECA region.
The same three impediments affect tertiary education in a slightly differ-
ent way, as this sector has already undergone significant reform over the
past 20 years, during which time it has doubled or tripled in size in many
adult education systems in the modern sense, these impediments cannot
be fairly characterized as problems of this sector, though the policy
recommendations offered in this chapter are applicable to adult educa-
tion also.
To be fair, ECA countries face these impediments to different degrees,
but none have fully escaped the legacy of central planning, which empha-
sized strict top-down control, compliance with pre-established norms, and
intensive management involvement in operational details.2 In terms of
measuring the quality and relevance of education, central planning focused
on generating data on inputs—that is, checking whether local actors were
in compliance with detailed norms for all inputs. Yet in terms of whether
students acquired skills and competencies, these systems operated with
the lights off—without the information needed—and as a result, educa-
tional spending in the region remains highly inefficient. Today, virtually no
ECA country has been able to downsize its school network and staff in the
face of falling student numbers. They are accordingly stuck with too many
schools and too many teachers, which absorb resources that could other-
wise be used to improve the quality of education.
Operating in the Dark: Ministries Know Too Little

to Effectively Manage the Education Sector
As mentioned in chapter 2, ministries of education across ECA know far
too little about the nature of the skills shortages in their economies or the
strengths and weaknesses of their own education systems. As a result, they
lack an important precondition for effectively managing the sector. Better
data alone does not make better policies. But operating in the darkness
makes it impossible to gauge the magnitude of the problems facing policy
makers. This section describes the current status of educational data and
policy making in the region, identifying the missing data needed and
showing how their absence impedes education in the region.
Resolving the Skills Shortage in the ECA Region: A Policy Framework 91
Educational Data Focus on Diplomas, not Skills

It is a problem worldwide that educational data tend to focus on
quantity—for example, the number of enrolled and graduating students—
and not the quality of education. Where international comparative data
are available, they focus on the quality of primary or lower secondary
education. Given that the current policy debate in the ECA region focuses
on “skills,” the lack of data on educational quality is problematic, particu-
larly because very few young people in the region enter the labor market
with only a lower secondary education. As figure 3.1 shows, most ECA
students complete at least an upper secondary degree. But with no inter-
national assessment of the skills and competencies of upper secondary or
tertiary graduates, it is impossible to quantify the gap in competencies
between recent labor market entrants in, say, Bulgaria and Germany. The
data that are available today document how many students graduate in a
particular year, not what they offer employers in terms of competencies.
Lack of relevant data on students and their individual performance is
particularly acute in the vocational sector because of the large variety of
vocational schools in the ECA region and the (likely) heterogeneity of
their student populations. The tendency of educational data to focus on
quantity, rather than quality and relevance, is also acute in adult education
Figure 3.1 Educational Background of 25–34-year-olds in the ECA Region, 2006

(percent)
100
6
90 17
25 26
80 40 31
70
60
56
50
61 61
40
30 57 63
20
27
10
14 12
0 3
EU10+1 MI CIS W. Balkans LI CIS Turkey
compulsory or less upper secondary beyond upper secondary
Source: Author’s calculations, based on EBRD-World Bank 2006.

and training. At best, current surveys in that sector measure the number
of training hours and courses in which individuals participate, or whether
a firm offers training. But no international comparative data yet exist to
compare the quality or relevance of such training.
Narrow focus and understanding of educational quality. Many ECA

countries tend to focus on their very best students and equate their
performance with “quality education.” Indeed, in discussions about the
quality of education generally, school principals, local education authori-
ties, and representatives of education ministries often point to particular
students’ accomplishments at the “Math Olympiads” as examples of how
well their schools perform. Similarly, university rectors mention how
many of their graduates manage to pursue further education at top uni-
versities abroad.
Although it is important to create an environment where excellence
can flourish, the few high-achieving students should not be the principal
focus of ECA school systems, nor the criteria by which they measure
their performance. Whereas the principals of ECA schools know how
many of their students compete in and win Math Olympiads, they know
very little about the average performance of their students, and much less
about the performance of their schools vis-à-vis similar schools elsewhere
in the country. As a result, policy makers have little information about the
performance of the weakest students in their schools: students with a dif-
ferent mother tongue, a different ethnic origin, or, simply, from economi-
cally disadvantaged backgrounds. These students are most in need of
support, yet, across the ECA region very little effort is made to under-
stand how far behind they are and whether existing policies—such as
small classes, extracurricular activities, or other interventions—are work-
ing. Without relevant data, problem areas go unnoticed, and the effective-
ness of policy interventions cannot be measured.
A broader understanding of educational performance relies on standard-
ized tests and looks at all students, differentiating their performance by
external factors (e.g., parental income and socioeconomic background). In
this view, truly outstanding schools are not necessarily those with the high-
est average score on standardized tests, but rather, schools that, compared
to their peers, manage to generate the biggest improvements in test scores.
Limited Participation in National and International Assessments

For some countries in the ECA region, information weaknesses are
compounded by limited participation in international student learning
assessments, such as PISA, PIRLS, and TIMSS. Five ECA countries

(Tajikistan, Turkmenistan, Uzbekistan, Belarus, and Kosovo) have never
participated in such an assessment. For these countries, it is impossible
to assess the quality of even primary and lower secondary education.
Moreover, three ECA countries have only just begun to participate in such
assessments.
ECA countries have also been slow to use national standardized assess-
ments to measure and analyze student learning outcomes, largely for
historical reasons. Prior to 1990, assessing students and granting diplomas
were left entirely in the hands of teachers and schools. A recent United
Nations Children’s Fund (UNICEF) report argues that this approach was
rooted in the philosophy “that external checks on outcomes were not
needed in a tightly controlled system of educational inputs and pro-
cesses.” As a result, “attempts in the early 1990s to introduce external
tests and examinations ran counter to the public mood, which saw them
as attempts to re-establish central control and undermine professional
trust in teachers’ judgment” (UNICEF 2007, 58).
Policy makers in ECA countries are recognizing, however, that
objective, standardized measures of learning outcomes are needed for
a number of reasons. First, they remain the best indicator available for
measuring the performance of an education system (Vegas and Petrow
2008). National and international learning assessments provide policy
makers a quantitative indicator of learning outcomes that can be
compared across schools and across time. These data in turn shed light
on which policies are working and which need tweaking; they also
illuminate weaknesses and identify areas where new policies may be
needed.
Second, standardized assessments allow students and their parents to
compare a student’s performance relative to his or her peers, and a
school’s performance relative to that of other schools. Third, when used
alongside other indicators of performance, standardized assessments can
be utilized to hold education managers and teachers accountable for
results. In a decentralized education system in which local authorities
play an important role in day-to-day management, such instruments are
especially important for a central government to identify system needs,
direct funding where it is most needed, and tailor new policies to support
schools and local authorities, as needed (see box 3.1).
Limited experience of using data to drive educational policy. Despite

some progress in developing national learning assessments and
Box 3.1
The Value of Standardized External Student Assessments

in ECA Countries
When small questionnaires on students’ socioeconomic backgrounds are
administered together with standardized tests (e.g., PIRLS, PISA, TIMSS—see
chapter 2), school systems can gauge their performance in imparting knowl-
edge and skills to their weakest students, who are often socioeconomically
disadvantaged. These students will never compete in Math Olympiads. They
drop out of school earlier than their peers, and, because they are marginal stu-
dents, never show up in national averages. Yet, an important strength of a school
system is measured by how many of these students—those with the weakest
parental support—acquire competencies needed for the labor market. Indeed,
the most successful school systems in the world, Finland and the Republic of
Korea among them, show that the way to raise overall educational quality for all
students is to focus on raising the educational quality provided to the lowest-
achieving part of the student population.
participating in international assessments, no country in the ECA region

has yet moved from regularly gathering data on learning outcomes
to using this data to improve policy making. The simple reason is that
each step along the path shown in figure 3.2 requires time and capacity
building. Specifically, moving from a culture where little or no data are
available to a culture where data are gathered, analyzed, disseminated,
and actively used to inform educational policies requires new skills.
On a positive note, some countries in the region (primarily the new
member states of the EU) now have several years’ experience with
performance measurement, both in terms of establishing their own
national assessment centers and participating in several rounds of
international assessments (see figure 3.3). This group of countries is
gradually adding the capacity to analyze test scores to design better
policies. However, this book finds that no ECA country has, as of 2010,
actually reached this stage. Less encouraging, a large number of coun-
tries in the region (mostly the low-income CIS countries, but also
Turkey and Belarus) have either made no or only small steps in this
direction (see far left column in figure 3.3). If this latter group of
Figure 3.2 Learning to Use Data to Drive Education Policy
assessing
learning
outcomes using
analysis to
identifying improve
employment analyzing and
disseminating policies
outcomes of
graduates data
e.g.,
performance-
collecting
based
data
budgeting
Source: Authors.
countries has begun to participate in international assessments, it has

only been quite recently.
No systematic tracking of graduates’ employment outcomes. When stu-

dents graduate, ministries of education in the ECA region do not system-
atically collect, analyze, and disseminate information on their employment
outcomes. Such information is useful for helping policy makers and
higher education institutions detect which programs and fields of study
are in high demand among employers. Moreover, such information can
help students make better choices about which university to attend and
which field of study to pursue.
Several OECD countries provide examples of how data on graduation
can be collected and disseminated. Norway has tracked such data since
1972; Italy, since 1998; and the Netherlands, since 1989. In the
Netherlands, almost all graduates of higher education institutions are
surveyed a year and a half after they graduate. The survey collects com-
prehensive information on a range of different topics, including informa-
tion on the school-to-work transition (asking such questions as, How long
did it take to find a job? Did it involve unemployment spells?); the type
and quality of employment, if any (e.g., sector and educational and skills
requirements); and students’ satisfaction with the education that they
have completed (Did it provide a solid basis for entering the labor mar-
ket? Did it develop the relevant skills? Did it achieve the right mix
Figure 3.3 Status of Measuring and Using Data on Student Learning Outcomes in
several years of
experience in
several years of measuring,
early piloting of participation in improve education
some
participation in Bulgaria
not started, very international
Hungary
early stages, no, or assessments
Albania
assessments Romania
Croatia
Armenia Serbia
Czech Republic
Azerbaijan Slovenia
Estonia
Belarus Georgia
Bosnia and Macedonia, FYR
Herzagovina Moldova
Kazakhstan Montenegro
Kosovo Poland
Kyrgyz Republic Russian Federation
Tajikistan Slovak Republic
Turkey Ukraine
Turkmenistan
Uzbekistan
in PISA, PIRLS, and TIMSS since 1995; and inputs from World Bank country experts. The UNICEF data is drawn from
“Table 2.1: Status of Reforms of Assessment and Examination Systems, 2006,” regarding “Introduction of other
school exams or assessments (e.g. basic school)” and “Introduction of sample-based national assessment.” In this
table, UNICEF scores the progress of countries on a range of 0 to 4, with 0 representing “not planned or started”
and 4 representing “operational.” In addition, the figure above uses World Bank staff compilations on the number
of international assessments in which each country has participated.
Note: As a signal that data-driven policy is a way to build capacity, the year in which countries first participated in
an international assessment was one of the indicators used to group countries.
between practical and theoretical knowledge?). Table 3.1 shows the types
of information that the survey collects.
Legacy of Central Planning

As noted earlier, many problems in education systems go unnoticed and
unaddressed in the ECA region because these systems do not produce
Table 3.1 Information Collected from Tracer Study of Dutch University Graduates, 2007
Duration of Full-time Unlimited Managerial or
job search employment term Monthly gross professional High use High job
(in months) (%) contract (%) income (euros) (% ISCO 1 or 2) of skills (%) satisfaction (%)
Mean Mean Mean Mean Median Mean Mean Mean
Science and math 0.7 88.4 45.6 2499 2429 83.6 71.0 77.5
Medicine and health 0.7 79.0 41.0 2904 2783 82.6 78.9 81.7
Engineering 1.0 94.4 66.2 2772 2631 87.4 75.2 72.8
Economics 1.0 96.1 71.3 2954 2783 70.3 65.6 70.1
Law 1.2 92.7 57.5 2864 2732 87.8 66.0 70.2
Humanities and arts 1.2 59.1 43.0 2188 2226 66.0 50.4 61.3
Social sciences 1.3 60.6 45.4 2317 2350 72.0 63.9 65.4
Agriculture 1.5 86.7 45.6 2137 2328 84.2 73.2 71.2
Source: Table provided by the Research Centre for Education and the Labour Market (ROA), The Netherlands, 2008, at request of the authors.
97
sufficient data to gauge their performance. The legacy of highly central-

ized, Soviet-style management then compounds these problems.3
Education authorities at most levels of the system have neither the power
nor the incentives to make key operating decisions, such as reallocating
budgets to meet urgent local needs, that would improve the performance
of their respective institutions (see next section). Instead, actors in the
sector are held accountable for complying with detailed, centrally deter-
mined norms (e.g. on class sizes, or the number of nonteaching staff to be
employed).
The implicit assumption appears to be that if norms are followed, a
quality education will result. There is little management understanding
that there are multiple ways to deliver quality education, some of which
must be discovered by the principals and teachers of individual schools.
Surveys of the main actors in the sector (undertaken as part of the PISA
and Teaching and Learning International Survey [TALIS] studies) verify
this general picture of education sector management in the region.4
Schools Lack Autonomy and Are Held Accountable

for Complying with Norms
ECA countries can be roughly sorted into three groups according to the
degree of autonomy granted to schools at the pre-university level. At one
end of the spectrum, schools in the new EU member states have been
granted significant autonomy over budgets, some aspects of staffing,
and—in some cases—instructional content. At the other end, schools in
low-income countries, such as the Kyrgyz Republic and Azerbaijan, enjoy
very little autonomy and continue to be micromanaged from the center.
Schools in the Russian Federation, Serbia, Turkey, and several other coun-
tries lie somewhere in between: some autonomy has been granted to
schools or to some schools as part of pilot projects.
Rather than provide a “framework” of broad, overall rules for the sec-
tor, legislation on education in the region is highly detailed, spelling out
minute norms for all operational areas (see box 3.2). Legions of inspec-
tors are employed to visit schools to verify if they are in compliance with
these norms, acting more like police officers than pedagogical counsel-
ors. The result is a system that focuses on compliance with norms, not
quality education.
Complying with norms does not necessarily lead to better results.

Detailed data on spending and test scores now exist for some ECA coun-
tries, including Poland, and show that complying with input standards
Box 3.2
Detailed Regulations Set Norms Even for Education

Facilities in the ECA Region
In Ukraine, the following school norms are set at the ministerial level. They offer an
example of the top-down management style that focuses on getting local actors
to comply with detailed norms, irrespective of whether or not they make sense in
local circumstances:
• 1 deputy school director for 11–24 class groups; 2 if more than 38 class groups
• 1 managing deputy director if an urban school has more than 600 students or
if a rural school has more than 400 students
• 1 pedagogue if a school has more than 8 class groups (regardless of the total
number of students in the school)
• 1 extracurricular activities group leader per each 16 class groups (regardless of
the total number of students in the school)
• 1 cleaner per 500 square meters (0.5 per each 250 square meters)
• 1 coat room attendant per 200 coat spots in school
• 1 yardkeeper per 1.5 hectares of school property
The examples above are commonplace across the region—remnants of

the types of detailed norms that governed all aspects of society during pre-
transition days. It is common for such norms to leave no room for a school or
local authority to consider, for example, alternative ways of achieving a clean
school (e.g., by outsourcing the task to a private company) or whether or not
staff specified in the norms are actually needed. Norms may even exist for which
there is no need, such as a norm that determines how many heat inspectors
should be hired per classroom, with no provision for whether or not classrooms
are actually in use.
Sources: World Bank 2008d; authors’ observations.
does not necessarily produce good-quality education. In fact, two very

similar localities can be in compliance with norms, but spend drastically
different amounts of resources per student and end up with the same
results. An analysis of educational expenditures and student test scores
in more than 3,000 municipalities of Poland that was undertaken for this
chapter shows that high-spending municipalities often end up with stu-
dents who perform worse on standardized tests than their peers in
municipalities that spend less.
An example of this analysis compares the two municipalities of Tarłów

(in the south-central region of Poland) and Rutka-Tartak (in the north-
east), both in compliance with norms for class sizes in Poland. The two
localities are also very similar in a number of ways: both are rural, have a
similar proportion of adults with secondary or higher education, and
roughly similar income levels. While they are similar in a number of ways,
the educational spending and results of the two towns vary widely—
suggesting large inefficiencies. Sixth-graders in Tarłów scored an average
of 18 out of 40 on a national achievement test in 2009, while sixth-
graders in Rutka-Tartak scored, on average, 26—a full eight points higher.
Oddly, the high-performing municipality, Rutka-Tartak, spent the least
per student: an average of only Zl 3,710 per pupil, compared to an aver-
age of Zl 8,330 per student in Tarłów (see figure 3.4). That is, Rutka-
Tartak achieved much better results at half the cost.
Further analysis revealed that the differences mostly boiled down to
differences in class sizes. Tarłów’s higher per student costs were a result
of relatively smaller classes (an average class size of 15), compared to an
average class size of 24 in Rutka-Tartak (see figure 3.5). Smaller classes
imply higher per student costs, yet test score results suggest that these
smaller classes are not resulting in better learning outcomes.
Figure 3.4 Relationship between Primary School Test Scores, Adult Education
Levels, and Per Student Spending in Two Municipalities of Poland
30
Rutka-Tartak
28 Zl 3,710
sixth-grade test score (max = 40)
26
24
22
20
18
16 Tarłów
Zl 8,330
14
10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 32.5 35.0 37.5 40.0 42.5 45.0 47.5 50.0
adults with secondary education or higher, as % of total
Source: Authors’ analysis of data from the Ministry of Finance, Central Statistical Office, Ministry of National Edu-
cation, and Central Examination Board, Poland, and the World Bank’s BOOST database; Kheyfets and others 2011.
Note: Size of bubble indicates relative size of per student spending.
Figure 3.5 Relationship between Primary School Test Scores, Average Class Size,
and Per Student Spending in Two Municipalities of Poland
30
Rutka-Tartak
28 Zl 3,710
sixth-grade test score (max = 40)
26
24
22
20
18
16
Tarłów
14 Zl 8,330
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
average class size
Source: Authors’ analysis of data from the Ministry of Finance, Central Statistical Office, Ministry of National Edu-
cation, and Central Examination Board, Poland, and the World Bank’s BOOST database; Kheyfets and others 2011.
Note: Size of bubble indicates relative size of per student spending. See end note for data sources used for this
analysis.
The story of Rutka-Tartak and Tarłów illustrates what is a widespread

phenomenon across the region: educational spending varies widely and
has little connection to learning outcomes. Furthermore, there are no
management mechanisms built into the education systems of these
countries to take corrective action in such cases. In fact, there are no
mechanisms to collect per student spending and test scores in different
localities. Not surprisingly, actors in these education systems care about
complying with norms, not about what it costs to deliver education ser-
vices or what results are achieved.
Inflexible Management Inhibits Reform

The strong focus on meeting detailed norms, together with the lack of
student performance data, has created rigid education systems that are
unable to respond to changing conditions. One area in which this
inability to reform is highly apparent is vocational education and train-
ing (VET). ECA education systems inherited large numbers of voca-
tional schools from the communist system, which in pre-transition days
supplied workers to state-owned enterprises. Despite large drops in
student numbers, these schools still enroll almost 40 percent of all
upper secondary graduates (the largest proportion of any region in the

world, see figure B2.1). The most highly vocationalized programs were
in Central and Eastern Europe, where the average share of vocational
secondary school students was over 70 percent in 1989. Turkey, not a
transition country, also has a strong vocational focus in secondary edu-
cation, with programs patterned after the German dual-system model.
As economic conditions and employer requirements have changed,
however, education systems in these countries have been unable to
reform the sector to deliver the skills in demand on the labor market.
In addition, the majority of these programs were terminal, that is, they
did not provide access to higher education.
Given that secondary education was overwhelmingly focused on voca-
tional and technical training in planned economies, the management of
vocational secondary education was often separated from that of general
secondary education programs. (The latter was the gateway to university
study, an opportunity that was limited to the select few in these econo-
mies.) In many ECA countries today, management of secondary educa-
tion is thus divided across two or more ministries. Even in Turkey, the one
nontransition country in the ECA region, there is a distinct fragmentation
of responsibility and oversight of secondary education within the same
ministry.
A further complication is that local governments in the ECA region
are typically responsible for the management and, to some degree, the
financing, of general secondary education in most ECA countries, whereas
the central government remains responsible for managing and funding
vocational secondary education. Specifically, in many countries it is the
national Ministry of Education that retained management control over
vocational schools (as did other national-level ministries that ran such
schools, for example, a Ministry of Agriculture that sponsored a forestry
school). The fragmentation is thus not only across ministries, but across
authority levels (national and local) also.
This double management fragmentation makes it more difficult to
adjust the supply of education programs to changes in demand (i.e.,
changes in skills requirements and labor market opportunities, which are
then translated into changing student demand). This reluctance or inca-
pacity to reform these education programs to meet skill needs is exacer-
bated by the fact that core ministries, including ministries of finance and
economy, are often more concerned than ministries of education—where
management control is actually located—about the efficiency of educa-
tion programs and their responsiveness to changing skill needs.
Due to the fractionalized authority structures, the lack of autonomy,

and a management focus on norms, principals of vocational schools or
local education planners have been virtually incapable of modernizing
these dinosaurs into attractive secondary education alternatives for
students. Economic transition necessitates widespread changes in educa-
tion at this level and in vocational education programs in particular, but
rigid and divided management systems and shrinking budgets have made
it hard to implement these changes. Many vocational schools remain
unreformed, with outdated equipment, an outdated curriculum that
offers too many narrow specializations, and an aging (and possibly
increasingly unqualified) teacher workforce. Paradoxically, firms in the
region are seeking more graduates with technical skills and have raised
salaries in areas with skills shortages to attract workers with these abilities.
Nevertheless, vocational programs in the region are still largely failing to
provide the needed skills.
Inefficient Use of Funds

Education funding in the ECA region is also dictated by norms rather
than a common-sense response to conditions on the ground. With money
tied to norms, schools and local authorities have no flexibility or incentive
to improve the efficiency of spending. In particular, schools in most coun-
tries in the region continue to be financed based on the number of classes
and teachers employed (i.e., inputs), not the number of students that
they serve. Centrally determined norms dictate the subjects that must be
taught, and how many classes must be created at each grade level. Norms
on class numbers determine the required number of teaching hours,
which establish the required number of teachers. Funding flows are then
rigidly earmarked for either personnel costs or nonpersonnel costs, leav-
ing local managers no ability to reallocate funds where they are most
needed. As a result, the system finds it difficult to correct course in the
face of change.
The most glaring example of this problem has been the inability of the
sector to downsize in response to falling student numbers over the past
two decades. The ensuing crisis has strained the resources of education
systems throughout the region and brought the teaching profession to the
brink—compromising both its quality and attractiveness to future teach-
ers. Although the problems associated with smaller student numbers are
presently most urgent at the pre-university level of education, they will
soon impact the university sector as well.
All but five ECA countries—Turkey and a few low-income CIS

countries—have experienced a dramatic decline in student enrollments as
a result of demographic trends in the past 20 years, differing only in terms
of the severity of the decline and its onset. As figure 3.6 shows, in some
countries student numbers have declined by as much as 40 percent since
the beginning of the transition.
Despite highly centralized management, the responsibility for identi-
fying which schools to close was placed on the shoulders of regional or
local authorities or both in most ECA countries. Yet these actors had no
incentive to undertake the politically sensitive and administratively dif-
ficult task of closing schools, particularly if their schools remained in
compliance with generous norms on class sizes. With norms for class sizes
allowing for variations as large as 25 percent (e.g., a minimum of 20 and
a maximum of 25 students) or with no minimum class size at all (e.g.,
Serbia and Belarus), years could pass before a school was not in compli-
ance with class size norms. Even then, local authorities usually had ways
to postpone closing schools by asking education ministries for exceptions,
adding additional financing from their own sources to keep schools in
operation, or inflating student numbers to keep schools in compliance.
Figure 3.6 Dynamics of the 6- to 12-year-old Population in the ECA Region,

1990–2006
(1990 = 100)
130
120
110
100
90
80
70
60
90
92
94
96
98
00
02
04
06
19
19
19
19
19
20
20
20
20
LI CIS MI CIS EU Western Balkana Turkey
Source: EdStats Database.

Note: LI = low-income, MI = Middle-income.
a. Western Balkan population estimates are based on Albania, Bosnia and Herzegovina, and FYR Macedonia only.
Plummeting Student-Teacher Ratios Reflect the Failure to Adjust

to New Realities
Declining student-teacher ratios across the region provide clear evidence
that policy adjustments to date have not successfully addressed inefficient
spending in the sector. Take the example of Moldova. The average school
in Moldova was built in 1971, when pre-university students totaled
1.2 million and demographics looked favorable. After the Soviet Union
collapsed, student numbers soon started to plummet. By 2010, there
were less than 700,000 students in the country’s schools (a 44 percent
drop from the peak student population of 1995). Yet the number of
teachers in these schools has not declined since 2003. In fact, there were
as many schools in Moldova in 2009 as there were in 1994; on average,
each school has only 56 percent of the students for which it was built.
Student-teacher ratios are one way to look at the inability to adjust to
falling student numbers. Before the transition, average student-teacher
ratios in the ECA region were roughly similar to those observed in other
regions in the world, but have since dropped significantly (see figure 3.7).
Figure 3.7 Primary School Student-Teacher Ratios in ECA Compared to Other

Regions of the World, 1990–2008
29
27
student-teacher ratio (primary)
25
23
21
19
17
15
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20

Source: EdStats database, plus author’s linear interpolations for missing years.
(in both the World Bank’s EdStats database and other international education databases) is that it is unclear
whether or not data for a particular country is reported on a full-time equivalent basis.
These low ratios are problematic for two reasons. First, other things
being equal, lower student-teacher ratios imply that it is more costly to
finance education per student. There is no evidence, however, that these
rising costs have resulted in better-quality education (for instance, see
figure 3.8 for the case of Romania, where TIMSS scores have remained
roughly flat for a decade). Second, lower student-teacher ratios mean that
more of the education sector’s scarce resources are tied up in employee
costs, leaving very few resources available for either innovations or much-
needed classroom learning materials (e.g., new textbooks, computers,
Internet access).
Teaching Profession Increasingly Devalued

The largest casualty of inefficient resource use has been the teaching pro-
fession, with evidence pointing to its significant devaluation—and a resul-
tant deterioration in its quality at the pre-university level. Given limited
resources, a teaching work force that is oversized relative to the number
of students is poorly paid; indeed, teachers in most ECA countries earn
salaries below the national median.5 Over time, low wages and uncer-
tainty about the future have diminished the attractiveness of teaching in
the ECA region.
The current teaching workforce has become overwhelmingly female
and is aging rapidly (see figure 3.9). The percentage of female primary
Figure 3.8 Real Per Student Expenditure Compared to TIMSS Math Scores in
Romania, 1999–2008
300 472 475 461 500

450
250
400
index, 1999 = 100
350
200
TIMSS score
300
150 250
200
100
150
100
50
50
0 0
99
00
01
02
03
04
05
06
07
08
19
20
20
20
20
20
20
20
20
20
TIMSS (math) expenditure per student (real, index 1999 = 100)
Sources: Authors’ calculations based on TIMSS (various years) and expenditure data obtained from Romania’s
Ministry of Public Finance.
Figure 3.9 Percentage of Students with a Teacher Over 50 Years Old in

ECA Countries, Selected Years
60
50
40
percent
30
20
10
0
Bulgaria
Hungary
Lithuania
Romania
Slovak Republic
Slovenia
Macedonia, FYR
Turkey
Russian
Federation
Armenia
EU10+1 SEE MI LI
CIS CIS
countries
1995 1999 2003 2007
Sources: IEA (1995, 1999, 2003, and 2007): Mathematics and Science Teacher Background Data Almanacs.
school teachers in non-EU members of the region is, in fact, considerably

higher than in the EU15 countries (EdStats Database). And education
systems throughout the region are finding it difficult to hire qualified
teachers in key subjects, such as English and information technology.
Finally, the quality of students now applying to pedagogical schools is
inferior to that of students applying to other programs of study, with few
of the brightest university graduates entering the profession. A recent
study (Silova 2009) looked at the situation of teacher training institu-
tions in two countries in the ECA region (Azerbaijan and the Kyrgyz
Republic) where detailed data on centralized university admission tests—
ideal for such analysis—were available. Consistent with the other indica-
tors reviewed in this section, Silova’s analysis supports the thesis that the
prestige and attractiveness of the teaching profession is low and declining
in both countries.
Considered separately, these facts might not be alarming. Taken
together, however, they make a compelling case that ECA countries are
failing to attract the professionals who are urgently needed to improve
the quality of education in the region.
Addressing the Skills Challenge

Education systems in the ECA region are facing daunting challenges with
limited resources. Resolution of the most urgent challenges of these
systems—overstaffing at the pre-university level, delivering the skills and
competencies needed on the contemporary job market, and revamping
vocational education and giving it an appropriate weight in secondary
education—can no longer be delayed.
The very urgency of these problems presents policy makers in the region
an opportunity to reform the way in which the education sector is man-
aged overall. Specifically, they can systematically collect student perfor-
mance data in order make better-informed decisions, make greater efforts
to reduce the role of the central government in day-to-day operational
issues at the school level, and link educational spending to agreed perfor-
mance goals. As will be discussed in more detail in chapter 5, some of these
principles have already been employed in many ECA countries at the ter-
tiary level. The challenge is to disseminate these same principles at all
levels of education.
Turn on the Lights

In order to provide quality learning for all, countries in the ECA region
first need to understand how much their students are actually learning. As
noted earlier in this chapter, all ECA countries need to begin systemati-
cally collecting data on how their education systems are performing,
principally via standardized student assessments and graduate tracer
studies (both of which need to be linked to student background informa-
tion). Central ministries can then use this data to design policy improve-
ments. To accommodate such a shift, education ministry staff will need
training in both analyzing trends in education performance data and
developing education policy options based on such data.
Participation in international assessments such as PISA and TIMSS is a
valuable step in this direction, as it gives countries access to significant
practical advice on how to design and administer student achievement
tests, analyze results, draw policy conclusions, and, equally important,
manage the dissemination and communication of these results to system
stakeholders and the media. This subject is discussed in more detail in the
following chapter.
Introduce Autonomy and Accountability at Lower Levels

of the Education System
Overcoming the legacy of central planning involves changing the focus of
education management away from detailed norms and instead holding
actors accountable for performance (or results). School principals and

local authorities need more decision-making authority to pursue new
opportunities, and the flexibility to experiment. At the central level, this
implies that policy makers in the ECA region would relinquish certain
duties and assume others so that education systems may innovate and
improve student learning outcomes. The opportunity before central gov-
ernments in these countries is to move away from micromanaging schools
and classrooms and focus instead on “steering” the system.
authority and responsibility in the hands of the people most able to inno-
vate and improve the quality of education: local managers and education
authorities (see Osborne and Gaebler 1992). The solution is not simply
to scrap norms altogether and replace them with expanded autonomy at
lower levels—a change in accountability relationships is also needed. The
art of reform will thus challenge central governments in ECA countries
to build new and more sophisticated relationships with local authorities,
principals, teachers, and other stakeholders—relationships based on
incentives and accountability relationships that make these actors part-
ners in achieving agreed performance goals. Not surprisingly, most sys-
temic education reforms that have taken place in OECD countries since
the 1980s have focused on devolving responsibility for day-to-day deci-
sions to the front lines, that is, to individual schools (OECD 2004b).
Aligning the incentives of stakeholders in the education sector with
the student learning outcomes desired by policy makers will require edu-
cation ministries to set overall performance goals; articulate who is
responsible and accountable to whom and for what; and ensure that these
responsibilities are agreed, accepted, and understood. Ministries will also
need to provide lower-level actors the support they need to meet agreed
education goals, whether in the form of financing, advice, or knowledge
sharing of best practices. Collaborative structures, shared leadership, and
the spirit of public service will be invaluable for supporting efforts to
improve learning outcomes in the region. Teachers, students, and local
education authorities need both resources and the organizational and
human capacity to perform. Teachers in the region, whose salaries are
rarely competitive with private sector wages, also need to feel valued,
respected, and recognized for their service.
Introduce Performance-based Financing

Education financing in the ECA region can be more than just a flow of
monetary resources, it can signal a desired policy direction to local educa-
tion managers. More flexible, smarter financing in the form of block
grants (i.e., contract- or performance-based) offers funding to learning

institutions in return for meeting agreed learning outcomes. Not only
does such financing give local education managers much-needed flexibility
that line-item, input-based budgets do not permit, it keeps them focused
on student results.
The majority of ECA countries have recognized the limitations of
input-based financing schemes and have come to appreciate that financ-
ing can be used as a policy instrument. At the pre-university level, a
typical first step away from input-based financing is to introduce a fund-
ing formula in which the number of students enrolled in a school is the
main component. In fact, 11 ECA countries have introduced such formu-
lae and fully abandoned input-based financing (see figure 3.10).6 In
essence, this means that public money is being allowed to “follow the
Figure 3.10 Progress Towards Results-based Education Financing in the

ECA Region, 2010
per student
financing widely
introduced
piloting of per
student financing
Armenia (2005)
Bulgaria (2008)
per student Kyrgyz Republic
financing under (2006) Czech Republic
discussion Moldova (2010) (1992)
Russian Federation Estonia (2001)
(1998) Georgia (2007)
Azerbaijan
not started or Tajikistan (2005) Hungary (1990)
Latvia
very early Uzbekistan (2008) Kosovo (2002)
stages in Macedonia, FYR
Lithuania (2001)
discussions Serbia
Poland (2000)
Slovenia
Albania Romania (2010)
Belarus Slovak Republic
Bosnia and (2004)
Herzegovina
Croatia
Kazakhstan
Montenegro
Turkey
Ukraine
Source: Authors’ assessment.

student” rather than inputs (e.g., staff). Money can follow the student in
many forms, such as student vouchers and block grants tied to student
numbers. Countries that introduce per student financing typically design
formulae that make it explicit and transparent how much a school will
receive (see chapter 4).
Summary
There are three major impediments preventing education systems in
ECA countries from helping to reduce the skills shortage in their region.
First, ECA education systems have been operating without crucial
information—in the dark. By focusing on top-achieving students, they
neither seek to deliver quality education to all students, nor do they sys-
tematically collect data on the learning and employment outcomes of all
students. Second, the legacy of central planning keeps the governance and
management of school systems highly centralized, with central policy
makers intensely involved in operational details. A focus on compliance
with norms, together with financing schemes based on inputs and not
outputs, means that most local education authorities and school princi-
pals in the ECA region lack the autonomy and authority to make crucial
management decisions for their own institutions.
Third, education systems in the ECA region use financial resources
highly inefficiently. Nowhere is this more apparent than in the pre-
university sector, where few countries have adjusted teacher staffing levels
in response to falling student numbers over the past 20 years. As a result,
student-teacher ratios have fallen sharply as per student costs have ris-
en—more so than in any other region in the world.
Policy makers in the region can overcome these impediments by com-
mitting to quality education for all students, supported by systematic data
collection on student outcomes; expanding autonomy at lower levels of
the education system in return for accountability for student perfor-
mance; and introducing performance-based financing. ECA countries are
specifically encouraged to develop national learning assessments, system-
atically participate in international assessments, conduct regular tracer
studies of graduates to determine their employment outcomes, and then
use this data to inform education policy.
Overcoming the legacy of central planning will require central policy
makers to reduce the role of the central government in day-to-day
operations of their education systems. By concentrating on setting system
standards and overall goals, the central government can then manage
education systems for performance instead of compliance with norms.

Greater autonomy can be granted to local education authorities and
school principals in return for their accountability for improved student
(and school) performance. Finally, policy makers can introduce perfor-
mance-based financing (through such mechanisms as block grants or
vouchers) and grant greater flexibility and decision-making power over
school budgets to local education managers in return for meeting agreed
learning outcomes.
The remainder of this report focuses on how governments can use the
conceptual framework presented in this chapter—the three key problems
impeding the education systems’ ability to address the skills shortage and
their corresponding policy solutions—to improve educational outcomes
in pre-university, tertiary, and adult education, respectively, with an
emphasis on practical options.
Notes
1. As a nontransition country, Turkey sometimes represents a different case.
Although much of the discussion in this chapter is relevant to Turkey, it will
be noted in those cases where it is not.
2. Although Turkey was not a command economy, it too has a very centralized
system, evident, for example, in the lack of autonomy and accountability in
the education system.
3. Again, though Turkey was not a command economy, it has a very centralized
education system.
4. The PISA assessment conducted by the OECD provides evidence of these
large differences in school autonomy, based on a questionnaire administered
to school principals (in which three questions relate to budgeting). Data from
the 2006 PISA show that schools in Estonia enjoy the most autonomy of all
ECA countries, irrespective of how the question on budgeting is asked, while
schools in Azerbaijan enjoy the least. It should be noted that PISA data must
be interpreted with some caution, as it is unclear what is meant by “having
more autonomy on budgeting.” Does it mean more autonomy to formulate
budgets, to re-allocate resources within budgets, to determine salaries and
annual increases, or something else? Still, certain ECA countries have unam-
biguously granted substantially more flexibility to schools, principals, and
governing boards than have others. A much better source of information on
differences in school autonomy is provided by the Teaching and Learning
International Survey (TALIS) study, also conducted by the OECD. In this
survey, teachers and principals are asked more specific questions related to the
degree of autonomy in their schools. Unfortunately, only six ECA countries
(Turkey, plus all new EU member states at the time) participated in the 2008
study, which confirmed that schools in Estonia enjoy substantial autonomy.
5. Comparing teacher salaries to those of other professions is not an easy task,
both because salary differentials can derive from differences in educational
background, experience, and other factors, and because different professions
enjoy different noncash benefits. However, available data show that teacher
salaries in the ECA region are low and that the gap between their salaries and
the median economy-wide salary has widened in many countries. See the fol-
lowing data sources on teacher salaries: for OECD countries, OECD 2008
(accessible online at www.oecd.org/edu/eag2010); for CIS countries, CISSTAT,
online database of the Interstate Commerce Committee of the Commonwealth
of Independent States (Statistics of the Countries of the CIS, Annual Data
from 1991 to 2010, http://www.cisstat.com/2base/frame00.htm); for economy-
wide annual salaries for all countries, OECD StatExtracts, http://stats.oecd
.org/Index.aspx. (All URLs accessed December 2010.)
6. The “result” referred to in this figure is a student enrolled. As discussed in
chapter 4, a more desirable per student financing scheme would be to finance
a student who is graduating, or better yet, a student who is graduating with a
desired level of competencies. However, even the most advanced ECA coun-
tries (in terms of moving to school financing based on results) still base
financing on the number of students enrolled (an intermediate result, at best).
This report therefore intentionally distinguishes between the inputs of teach-
ers, classes, and finances, and among students “enrolled,” “graduating,” “gradu-
ating with a desired level of competencies,” or “graduating and finding a job”
as results (or outputs).
CHAPTER 4
Managing for Results at the

Pre-University Level of Education
The policy framework elaborated in chapter 3 can be applied to pri-

mary and secondary education in the ECA region to help create edu-
cation systems more capable of meeting the skills gap outlined in
chapter 1. First, policy makers in the region can overcome the debili-
tating lack of data on the performance of their education systems.
Second, a more sophisticated relationship can be put in place between
the central government and lower-level actors in the school system
based on this data: greater autonomy at lower levels of the school
system in return for holding local educational authorities and school
directors accountable for improved student performance.
Freed from the day-to-day management of schools, policy makers will
be able to devote more of their attention to elaborating strategic educa-
tional policy, including research and analysis of system and fiscal perfor-
mance, and setting overall system goals, standards, regulations, and
guidelines. One area of immediate concern will be designing long-term
teacher policies to renew the teaching force and re-establish the prestige
of the teaching profession in the ECA region.
Finally, current sector resources could likely achieve superior educa-
tional outcomes if schools were larger and if fewer but better paid teach-
ers taught slightly larger classes. Linking financing directly to student
115
performance will help to increase the efficiency of fiscal resources and

hold local-level actors accountable for achieving the education goals of
the central government. Special care must, however, be taken to prevent
performance goals and financing mechanisms from punishing schools
that have larger numbers of lesser-achieving students and instead reward
them for improving these students’ academic achievement. In the end,
financing modalities that give schools the flexibility to allocate resources
to the areas of greatest need will help schools improve the performance
of all students, not just the highest achievers.
Track Student Learning and Employment Outcomes

Better data alone do not lead to better-quality education. Partly due to
difficult demographic trends and partly due to history, ECA education
systems have wound themselves in a knot that will be very difficult to
untie without devoting greater attention to data on results. The legacy of
central planning has created a system where bureaucrats manage the sec-
tor by writing detailed norms into legislation and then managing princi-
pals so that they comply with these norms. Generations of managers have
lived with this system for their entire lives. When asked to improve edu-
cation system performance, they therefore reach for the tools that they
have always used, either by revising norms or ramping up inspections.
This chapter argues that the solution is not to rewrite norms or expand
the inspectorate but rather, to reach for different tools. However, none
of the new tools will work without better data on performance: what
competencies are students mastering and are graduates finding jobs?
As described in chapter 3, management of the education sector needs
to shift from managing inputs to managing for results. This type of man-
agement requires a more sophisticated partnership between the central
government and actors at the school (and local government) level, based
on a shared and measurable understanding of what is to be achieved.
That is, data on performance are what enables the central government to
start managing for results, rather than for inputs. The chief source of this
performance data is national and international student learning assess-
ments, together with surveys of graduates that determine their employ-
ment outcomes.
Implement Yearly National Assessments

As noted earlier in this report, many ECA countries have begun to develop
and, in some cases, implement national assessment exams. However, few
Managing for Results at the Pre-University Level of Education 117
(if any) countries in ECA are yet using the results of these exams to help
design educational policies. To do so effectively, several key actions are
needed. Most importantly, some background information about the stu-
dents who take national assessment exams needs to be collected. The
importance of this information should not be underemphasized: in its
absence, the usefulness of test scores is severely constrained.
In Bulgaria, for example, a national assessment center can generate
average national and regional scores on student achievement tests, but
this data cannot be disaggregated to identify the performance of either
different groups of students or individual students. The problem is that
the assessment system (which gathers data on student test scores) is not
linked to either socioeconomic information about students (e.g., data
collected via survey when students take the test) or to the individual
student identification numbers assigned by the Ministry’s education
information system. This oversight means that education policy makers
in Bulgaria are unable to answer crucial questions related to the large
wave of school closures that took place during the summers of 2007–
2009, such as whether students who were moved to new schools are
doing better or worse academically.
Further recommendations to improve assessment systems include the
need to administer such tests on an annual basis, not only to collect stan-
dardized longitudinal data, but also to enable countries to have a baseline
of student achievement results before and after educational reforms are
introduced. Moreover, countries need to move beyond using national
assessments solely for examination purposes (that is, to certify that a stu-
dent has mastered a curriculum at the end of compulsory and upper
secondary education) and begin using such tests to assess the quality of
education at different stages in the cycle. For instance, policy makers may
want to use a standardized test administered each year in order to evalu-
ate how well fourth-graders are reading.
Continue (or Initiate) Regular Participation in

International Assessments
International assessments give countries outside benchmarks with which
they can gauge how well their education systems are performing. In addi-
tion to being able to track their performance against countries worldwide,
participation in such tests as PIRLS, TIMSS, and PISA, will give ECA
countries access to substantial technical assistance in how to analyze,
understand, and use test results to design better education policies. Finally,
systematic participation in international assessments enable countries to
Box 4.1
Using Data to Measure Gaps and Design Better Policies:

Three Examples
In Serbia, researchers used PISA test scores for 2006 and 2009, together with student
socioeconomic and ethnic information, to shed more light on what is widely known
in the region: Roma students significantly lag their non-Roma peers (see table B4.1).
Through careful analysis of this data, the researchers were also able to show that
part of this performance gap was due to socioeconomic factors. That is, students
with a Roma background were, on average, from more disadvantaged backgrounds
(such students tend to do worse than their peers from better-off families). However,
even when controlling for the fact that Roma students tend to come from more
disadvantaged backgrounds, the gap is large, especially in math, suggesting that
Roma students will need additional support to catch up, over and above that need-
ed by children from other socioeconomically weak backgrounds. Such information
on its own does not provide policy recommendations, but it sheds light on where
problems exist, quantifies them, and provides a baseline against which policy initia-
tives can be measured.
Table B4.1 Learning Gaps between Roma and Non-Roma Students in

Serbia, as Measured by PISA 2006 and 2009 Results
PISA 2006 PISA 2009
Gap without Gap when Gap without Gap when
controlling for controlling for controlling for controlling for
socioeconomic socioeconomic socioeconomic socioeconomic
differences differences differences differences
Matha 107 77 86 43
Reading 77 44 91 52
Science 81 45 81 38
Source: Baucal and Pavlović-Babić 2010. These findings are corroborated by Baucal’s 2006 study of the
performance of Roma children on Serbia’s national school assessment exam.
Note: Gaps shown above are calculated as the score of non-Roma students minus the score of Roma
students.
a Statistically significant.
In Romania, researchers used PIRLS 2001 and 2006 data to quantify the extent
to which students from rural schools lagged behind their peers from urban schools
(Romania 2009). Equally important, they pointed out that the gap had widened, not
narrowed, between the two tests. The magnitude of the gap played an important
role in triggering the design of a project (financed by the World Bank) to address the
quality of education in rural areas (see World Bank 2003b).
Box 4.1 (continued)
Finally, to gauge the risk that tertiary students were withdrawing from school
as a result of the ongoing financial crisis, researchers in Romania administered a
survey of students to identify those most at risk of dropping out (Pricopie et al.
2010). These results can be used by policy makers to design policies to mitigate
this risk.
examine their educational progress over time because the tests are com-
patible and administered at pre-determined time intervals (e.g., PISA is
administered every three years; PIRLS, every five years).
Monitor Student Labor Market Outcomes

The labor market outcomes of secondary graduates is another critical part
of the performance data needed to manage education systems for results.
This information is particularly important for transforming secondary
education in the ECA region. This level of education can only respond to
student and employers’ demand if policy makers have accurate and
timely information about the career implications of students’ educational
choices. In principle, this type of data should be gathered through regular
labor market surveys (which exist in most ECA countries), periodic
employer surveys (which are implemented in most countries on an ad-
hoc basis), and graduate tracer studies (which exist nowhere in the
region, save Romania and Hungary as of 2011). Tellingly, most countries
have statistical agencies that conduct regular labor force surveys, but
ministries of education tend not to analyze and use such results to design
a more responsive education sector. Moreover, the results of these surveys
are rarely made available to students or parents in a readily digestible and
user-friendly form.
Create demand for more data and analysis. In order to mobilize and
sustain resources to collect and analyze data, demand for this data and
analysis must be stimulated. If such data are not used in the policy process,
moreover, policy makers are quick to axe resources allocated for these
purposes. Given the weak internal capacity of most ministries of educa-
tion in the region, central governments might consider relying on outside
researchers—such as university and policy institute experts—for help. This
could be done in a number of different ways, including the following:
• Set aside competitive grants for researchers to write policy notes and
research papers using assessment data.
• Make entire datasets more readily available. For instance, the OECD
makes the entire PISA dataset publicly available for researchers to
download online (http://pisa2009.acer.edu.au/; accessed December
2010).
• Change policies to make data matter, such as making improved student
learning outcomes a component of teacher and school evaluations. (If
the results of assessments have no impact, whether directly or indi-
rectly, on policies or people’s careers, it is very difficult to mobilize and
sustain resources to maintain assessment centers and train people to
analyze results.)
Expand Autonomy in Exchange for Accountability for Results

The OECD (2004b) has recognized that high-performing education sys-
tems tend to have local schools and education authorities with a high
degree of autonomy. Not surprisingly, most systemic education reforms
that have taken place in OECD countries since the 1980s have focused
on devolving responsibility for day-to-day decisions to the front lines, that
is, to individual schools. The underlying logic behind such devolution is
to empower school principals. These principals, who are familiar with
their respective staff, students’ needs, and local conditions, are better
suited to make operational decisions than bureaucrats in capital cities.
School-based management, moreover, allows for stronger accountability
relationships than micromanagement from the center (see, for instance,
Barrera-Osorio, Fasih, and Patrinos 2009).
Policy makers can employ a range of different options for strengthen-
ing accountability, some of which are more rudimentary—and can be put
in place faster—than others. All of these options entail risks and trade-offs
that must be considered to prevent unwanted effects. With respect to
school performance, ministries of education can do the following:
• Create and empower school councils and create incentives for their
active participation in school decision-making processes, and provide
them with adequate capacity and training, to increase the participation
of parents and the community (see Patrinos et al. 2010).
• Administer national student assessments at key stages of the education
cycle (e.g., after the 4th, 7th, and 12th grades) and make this data read-
ily available to parents and school councils to help hold schools ac-
countable for results. Making assessment results more readily available
has helped improve state-level accountability and student achievement
in the United States (Hanushek and Raymond 2005; Carnoy and Loeb
2002) and Mexico (Alvarez et al. 2007).
• Require schools to prepare school development plans that outline each
school’s strengths and weaknesses, together with a vision and action
plan for making improvements. Ideally this exercise should take as its
starting point student scores on a national assessment exam, together
with an institutional self-evaluation.
• Prepare “school scorecards” for all schools that include basic indicators
(whichever are available), giving parents and students a sense of a
school’s performance. Such scorecards should be made available on the
ministry’s website and disseminated to local authorities and schools.
Such score cards should publish the “value-added” of a school (e.g., an
increase in test scores that takes into consideration students’ socioeco-
nomic background).1
• (Ideally) disaggregate school score cards—where regulations permit—
by the performance of
° Minority students
° Low-income students
° Students with limited proficiency in the language of instruction
° Students with special education needs or disabilities or both
• Agree on criteria that objectively identify “a school in need of improve-
ment” and decide what actions will be taken to support such schools.
More specifically, they can agree on the ultimate sanctions for contin-
ued failure to improve performance, such as
° Re-constituting the school (i.e., formally closing it down but reopen-
ing it immediately with new management and the possibility of staff
re-applying for their jobs).
° Turning the school over to private management
• Provide rewards to high-performing or improving schools or both. Here
policy makers should avoid characterizing a “high-performing school”
simply as one that has high average student scores on a standardized
test or a large number of graduates who go on to university, as these
indicators may merely reflect the fact that the school attracts high-
caliber students (possibly from well-off families). Rather, a school
should be judged as high- (low-) performing based on whether it deliv-

ers (does not deliver) learning to all types of students, including those
from low-income and minority households. This is why it is so impor-
tant that the performance of a school, in terms of student learning
outcomes, be disaggregated to show the achievement of different types
of students (e.g., low-income and minority students).
At the level of the school, education ministries can do as follows:
• Require that principals provide explanations for a schools’ failure to

improve student learning outcomes.
• Revisit the issue of who can hire and fire a school principal and to
whom he or she is accountable. Given enhanced autonomy and more
flexible financing, the most important question to ask is: Who is best
positioned to ensure that a principal performs? The education minis-
try? A regional inspector? The local mayor? A school council (perhaps
with representatives of all of the above)?
• Require that school principals undertake training and be licensed and
institute programs to make this possible.
There are numerous risks related to holding schools accountable for

results that policy makers must acknowledge and manage. For example,
if an accountability system is based on tests that measure basic skills—
and if these “high-stakes” tests carry great weight in school evaluations—
the system risks focusing the education system solely on the measurement
of basic skills, ignoring the overall goal of creating well-rounded, cre-
ative, active citizens and productive workers capable of engaging in
lifetime learning.
A well-functioning accountability system requires credible consequences
for success and poor performance, recognizing that outside factors also play
a role in determining student learning outcomes. In the case of success,
schools need to be recognized and rewarded. In the case of poor perfor-
mance, schools need to be sanctioned in some way. At the very least, if
learning outcomes deteriorate at a given school, the principal should pro-
vide an explanation for why this has happened (e.g., the composition of the
student body changed in such a way that the school now has more students
from challenging socioeconomic backgrounds). In the United States, the
ultimate sanction is to “reconstitute” a school, which means replacing the
principal and asking all teachers to re-apply for their jobs. In most U.S.
states that have the power to reconstitute schools, however, schools are first
placed on probation for two to four years, during which time they receive
support and guidance to improve student achievement.2
Relax Class Size Norms

In the ECA region, moving toward greater school autonomy implies
relaxing or abandoning the detailed norms or both that govern virtually
all operational decisions. The most crippling norm that constrains school
principals from managing for results is the class size norm. As noted in
chapter 3, this norm—together with a centrally imposed curriculum—
determines how many teachers can be hired. This number in turn deter-
mines 65 percent or more of most schools’ annual budgets. Thus, this
norm, more than any other, restricts principals from making the best deci-
sions for their schools in terms of student learning and financial efficiency.
The regulation is, moreover, fiercely protected by teacher’s labor unions
because it protects jobs. In this respect, Denmark provides an example of
how to gradually eliminate class size norms (see box 4.2).
Abandon the Policing Mindset

Moving toward greater autonomy also implies changing a “policing”
mindset that treats all schools more or less the same to an “auditing”
mindset that differentiates management based on local capacity. Today,
many schools in ECA countries are “policed” by inspectors, Ministry of
Education staff members and financial managers: every school is visited
at the beginning and end of a school year to check if norms and regula-
tions are being complied with. Few schools have their own bank
accounts and those that do sometimes need multiple approvals to make
mid-year corrections to their initial budgets. The entire system seems
designed based on the implicit assumption that wrong-doing can only
be prevented by wrapping lower-level actors in as many layers of control
as possible.
This “policing” culture will need to be abandoned to make local or
school-based autonomy effective. Ministry of Education staff mem-
bers and inspectors may very well need to continue to vigilantly man-
age certain schools (e.g., small schools with poor capacity), but others
(e.g., large, well-established, and well-run schools) can be managed
with a much lighter hand. For instance, rather than making every
transaction subject to approval (still the case in many countries), well-
run schools should be evaluated on the basis of performance criteria
(i.e., specific learning outcomes) and year-end auditing of their finan-
cial statements.
Box 4.2
Eliminating Class Size Norms in Denmark

Upper secondary schools in Denmark used to be governed by class size norms
similar to those that exist across the ECA region today. Today, however, school
principals have the discretion to determine the appropriate class size, while
ensuring that their schools remain attractive in the eyes of prospective students.
The Ministry of Education in Denmark long mandated that class sizes in upper
secondary schools could not exceed 24 students. This norm was removed by the
late 1960s, after which labor contracts with teacher unions specified that teachers
teaching classes of more than 28 students would receive (substantial) additional
compensation. Since 1999, an agreement was reached between the unions and
the central ministry to abandon an upper limit, allowing school principals to de-
termine this limit.a
In determining the appropriate class size, principals must weigh economic,
pedagogical, and “marketing” factors. Since schools in Denmark are considered
self-owned institutions, economic factors are an important consideration. Col-
lapsing two classes into one allows for considerable savings. On the other hand,
principals must take into account pedagogical considerations so as to ensure
high-quality education. Finally, principals know that students and parents—who
are free to choose the school that their children will attend—value both smaller
classes and a large selection of free electives.
Source: Meeting with school administrator in Denmark, summer 2008.
Note: a. Certain broad guidelines and requirements exist regarding class sizes. For instance, if 10 students
at a particular school choose a particular “free elective,” the school must offer it. Moreover, the ministry
can mandate that a particular school offer a particular course in the interest of making sure that a specific
geographic area of Denmark has sufficiently broad course offerings.
The ultimate goal of greater autonomy is to make an education sys-

tem more flexible and responsive to student and employer demands.
This is particularly needed in vocational education and training in the
ECA region (see box 4.3), enabling it to expand or contract course
offerings in response to student and employer demand, not on rigid
governmental norms.
Introduce Autonomy Gradually to Offset Corruption

Education and finance officials in the region are rightly concerned about
expanding institutional autonomy in an environment characterized by
weak managerial capacity and widespread corruption. Such concerns are
justified: as chapter 5 points out, the experience of expanding institutional
Box 4.3
Rethinking Secondary Education

One area where more local flexibility is needed is in the area of vocational education,
where ministries of education in the ECA region continue to play too dominant a
role. In many countries, these ministries continue to directly manage the profile,
course offerings, and budgets of vocational schools. This has left the sector largely
underfunded and unreformed. It has also maintained a strong dichotomy between
vocational and general education schools. This chapter recommends expanding
autonomy in this field by allowing either principals or local authorities (whether at
the municipal or regional level) a greater say in determining the future of their
schools, irrespective of their profile. In exchange for more autonomy, school princi-
pals should be held accountable for improvements in learning outcomes or
improvements in the rate at which graduates progress to further study or find jobs.
The still strong dichotomy between vocational and general education in the
ECA region stands in contrast to the trend worldwide, which is away from separat-
ing secondary education into general and vocational tracks and toward a com-
mon secondary curriculum. This curriculum offers students course options and
defers job-specific education and training until after secondary school (Rutkowski
and Scarpetta 2005). Moreover, many of the skills for which worldwide demand is
growing most rapidly—such as foreign-languages and scientific problem-solving
proficiencies—are normally provided by general secondary and higher education
programs. To that extent, general secondary education is becoming, in effect, rel-
evant vocational education. Given the need for lifelong learning and skills upgrad-
ing in a global knowledge economy, all secondary education programs should
thus aim to provide the foundation for further educational development, includ-
ing the deepening and refinement of skills.
Recent reforms in Poland, which led to consistent improvement in student
learning outcomes, provide a powerful example of how educational quality at the
secondary level can be enhanced by postponing the early sorting and streaming
of students into academic and vocational tracks. Research suggests that the im-
provement in Polish students’ test scores can be traced to the increased exposure
of all students, especially low-performing students, to academic instruction
(Jakubowski et al. 2010). This goal was achieved by delaying the start of voca-
tional education, with its traditionally greater emphasis on narrow job-related
skills, by one year for all students. The result was much stronger grounding of
students in reading and mathematics, especially among weaker students who
had previously been tracked into vocational schools, and consequently, better
performance in assessments of student learning such as the PISA.
Sources: Rutkowski and Scarpetta 2005; Jakubowksi et al. 2010; authors’ analysis.
autonomy at the tertiary level in the ECA region has not been without
problems. Unfortunately, unethical behavior is flourishing at the tertiary
level, which undermines the sector’s ability to deliver quality education
for all.
There are, however, ways of moving forward at the pre-university level
while addressing these concerns. For instance, institutional autonomy
does not have to be granted to all schools at the same time. Since central
authorities might reasonably have concerns about granting autonomy for
management when capacity at lower levels is very limited, one practical
approach would be to start with larger schools (and larger local educa-
tional authorities), which usually have more managerial experience. From
an implementation perspective, the benefits of focusing on large schools
are obvious. In the case of Romania, for example, focusing initially only
on large schools would involve training, say, 1,000 school principals and
accountants, rather than the principals and accountants of more than
7,000 small schools.
The degree of autonomy granted to schools can also be expanded
gradually, with country circumstances influencing the ultimate degree of
autonomy eventually granted. It is important to note that autonomy—
like accountability—is complicated. For instance, granting schools more
autonomy in budgetary matters is unlikely to have much impact unless
this autonomy includes the ability to make staffing decisions (since staff
costs are the single largest budget item). Similarly, autonomy over staffing
decisions will, inevitably, involve providing schools the autonomy to
determine class sizes. Policy makers are thus urged to simultaneously
make changes to a number of rules and regulations as part of a change in
class sizes.
Improve the Efficiency of Resource Use

An essential step toward more results-oriented management of basic
education is linking funding to results, which will advance several objec-
tives. First, performance-based funding signals the desired result. Because
information on results enters budget discussions, education systems are
forced to collect data and measure results more systematically, creating
“demand” for student achievement data and its analysis. Second, such
funding embeds incentives into financing flows: actors in the system are
rewarded by delivering results and will therefore seek to do so. As a
result, performance-based financing can help address the large inefficien-
cies characteristic of education sectors in the ECA region. With money

linked to results—as opposed to inputs—principals will, for example, hire
the number of teachers and janitors needed (and that can be afforded),
not the number mandated by norms.
In addition to instituting performance-based financing, education min-
istries can do the following:
• Require local authorities and schools to present their expenditures in a

specific format that allows them to be compared nationwide, as well as
have them audited by an outside party.
• Require schools to run surpluses or deficits no greater than a specified
percentage of their total budget for a specified time period, say, one
year.
• Require that school budgets and actual expenditures be made publicly
available on school bulletin boards or websites or both.
Per Student Financing

In pre-university education, the most basic way to introduce perfor-
mance-based financing—the path that several ECA countries have begun
to pursue (see figure 3.10)—is to tie school funding to the number of
students enrolled. Hence this type of financing is referred to as “per
capita financing” or “per student financing.” Theoretically, defining a “per-
formance” measure as a student “enrolled” is somewhat unappealing.
A theoretically more appealing measure would be “a graduate” or even
“a graduate who has obtained a certain level of proficiency” (or even “a
graduate who finds a job within a fixed time frame”).
It is important to note that per student financing does not imply that
the number of students enrolled is the only factor that determines the
amount of money transferred from the central government to a local
school. In fact, most countries allow the per student amount to vary
according to a formula. Allowing for different per student amounts makes
sense because the cost of providing education is not uniform in a given
country. For example, it is more costly to enroll and graduate a socioeco-
nomically disadvantaged or minority student because this student will
require more attention and support. Similarly, areas of the country with
few inhabitants will naturally have smaller schools that, in turn, imply
larger costs. Good formulae take these differences into account to avoid
punishing schools and local authorities that serve challenging student
populations or locations or both. Box 4.4 provides an example of a
Box 4.4
Per Student Funding Formula: Recognizing the Varied Costs

of Providing Education
The new financing system that became effective in Bulgaria on January 1, 2007,
establishes individual per student financing standards for four different types of
municipalities. Figure B4.4 presents these per student standards and the criteria that
were used to differentiate between municipalities. The first type consists of munici-
palities that have 70,000 or more people living in the municipal center. The remain-
ing municipalities (all of which have less than 70,000 inhabitants living in the
municipal center) are divided into three groups: (1) municipalities with a population
density greater than (or equal to) 65 people per square kilometer; (2) municipalities
with a population density less than 65 people per square kilometer; and (3) small
mountainous municipalities (a subgroup of 1 and 2) with more than three settle-
ments and less than 10,000 inhabitants living in their municipal centers.
Whenever the financing scheme for education is changed, there are winners
and losers. Identifying the losers and providing them with an adjustment path is
thus part of a good implementation program. In Bulgaria’s case, the Ministry of
Education and Science (MES) recognized that 88 out of 264 municipalities would
Figure B4.4 Groupings of Bulgarian Municipalities, Together with Baseline

Per Student Financing Amounts and Adjustment Coefficients, 2007
large municipalities: underpopulated

population ≥ 70,000 in the 40 municipalities:
municipalities:
municipal center baseline amount x 1.07,
population < 70,000
or Lev 849
in the municipal
70 center and a population
municipalities: density ≥ 65 persons
15 baseline per square meter
municipalities: amount x 1.2,
baseline or Lev 958
139 municipalities:
baseline amount x 1.12,
or Lev 894
other municipalities: small mountainous
population < 70,000 in municipalities:
the municipal center and population < 10,000 in the
population density < 65 municipal center and ≥ 3
persons per square meter separate settlements
Source: World Bank 2007b.

Box 4.4 (continued)
receive less funding (in nominal terms) under the per capita scheme than they
had received the year before the new scheme was introduced (i.e., in 2006). To
allow these municipalities time to adjust their class sizes (and, if needed, reduce
the number of schools), they were entitled to receive compensation equivalent
to the difference between the allocation based on the new per student amount
and their allocation in 2006 during the initial year of implementation.a
However, to ensure that these “losing” municipalities would eventually be able
to operate in a fiscally sustainably manner under the new regime, they were
required to present a school consolidation plan to MES and the Ministry of Finance
in order to qualify for the compensation. In addition, the MES clearly communi-
cated that the compensation would only be available during the first year of im-
plementation. Thus, losing municipalities had strong incentives to initiate school
consolidation; with wages increasing by about 5–10 percent per year (at the time)
in nominal terms, the only way for them to become financially viable under the
per capita scheme was to consolidate their school networks. As expected, the
reforms led to a large number of school closures, as municipalities moved to cre-
ate more viable schools.
Source: Authors based on World Bank 2007b.
a. On the other hand, implementation plans did not specify any ceilings on funding for municipalities
that would be “winners” under the new scheme. Thus, municipalities that had well-optimized school
systems would experience increases in their education budgets.
formula used in Bulgaria that differentiates between the underlying costs

of providing education in different circumstances.
Create incentives for larger class sizes. As mentioned in chapters 2 and

3, student-teacher ratios, class sizes, and school sizes have fallen precipi-
tously in most ECA countries over the past 20 to 30 years. As a result,
education systems in most of these countries are wasting scarce resources
(from an educational perspective): they are heating and maintaining half-
empty buildings; employing, training, and paying salaries to a larger staff
than needed; and maintaining class sizes that could, in all likelihood, be
increased (e.g., to their levels of just 20 years ago) without affecting the
quality of education.
Small class sizes and a large number of teachers relative to students
are observed in all countries of the region for which data are available,
except Turkey. For instance, class sizes at the primary level (i.e., grades
1–4) in Poland, Serbia, and Azerbaijan are 20, 14, and 12, respectively,
compared to 23 in France, 26 in the United Kingdom, and 31 in Chile

and the Republic of Korea (see figure 4.1). School-level data show that
the small class problem in the ECA region is not only a problem of small
schools—even the largest schools in these countries have relatively
small classes (see figure 4.2). The fact that even large schools have small
class sizes suggests that significant savings can be generated merely by
within-school optimization, that is, forcing existing (large) schools to cre-
ate larger classes.
As noted in chapter 3, there is no “magic” size that fits all students and
learning circumstances. What works, on average, in a Korean school in
2010 may not work in a Ukranian school in 2010, and what works for
some students may not work for others. A class size of 35 may be appropri-
ate in an urban classroom taught by the country’s best teachers, whereas a
class size of 5 may be suitable for a tiny, rural school where every student
is from an ethnic minority. Rather than focus on determining a general
“appropriate” class, ministries of education would thus be better off trying
to change the rules of the game for principals and local authorities.
Specifically, as discussed in the previous section, they should expand local
autonomy to let principals determine the appropriate class size for their
respective schools (taking into consideration local circumstances and
needs) and use a per student financing formula to ensure that financial
performance is also part of a principal’s management considerations.
Encourage school optimization. The experience of ECA countries that

have implemented per student formulae suggests that these formulae
alone do not lead to school closures; their introduction needs to be
complemented by other policy initiatives (e.g., a program to finance the
transportation of children to their new schools). However, putting such a
formula in place can signal to school principals and local communities
that certain schools are no longer financially viable.
Until its recent push to close underutilized schools, Bulgaria had
many villages with large school buildings capable of accommodating
hundreds of students (see box 4.5). These schools were standing virtu-
ally empty, with only 50 to 100 students enrolled. Addressing the
inefficiencies that result from such an oversized school network—
underutilized space, small class sizes, and scarce maintenance resources
stretched across many buildings—is a necessity throughout the ECA
region. Closing and merging schools will, however, be substantially
more difficult than increasing class sizes in large schools. In addition,
school closures may produce certain associated costs, including
Figure 4.1 Comparison of Average Size of Primary School Classes Worldwide, Various Years
35
30
25
20
15
10
0
Slovenia
Estonia
Slovak Republic
Poland
Hungary
Czech Republic
Bulgaria
Azerbaijan
Armenia
Ukraine
Russian Federation
Turkey
Serbia
Kosovo
Luxembourg
Greece
Spain
Finland
Germany
France
Ireland
United Kingdom
Mexico
United States
Brazil
Chile
Korea
OECD Average
EU10+1 LI CIS MI CIS Turkey Balkans Western Europe other
Sources: OECD Education At a Glance Database for OECD and partner countries; authors’ calculations based on Ministry of Education data (from education management information
systems) for Azerbaijan, Serbia, and Romania, and on National Statistical Institute school-level data for Bulgaria.
Note: Most recent data for each country (2005, 2006, 2007, or 2008).
131
Figure 4.2 Average Class Size in Large Primary and Secondary Schools in Nine
ECA Countries, Various Years
35
average class size (number of students)
30
25
20
15
10
0
9)
8)
7)
9)
9)
9)
9)
7)
10
00
00
00
00
00
00
00
00
20
(2
(2
(2
(2
(2
(2
(2
(2
a(
va
ia
ia
ia
ia
ia
ni
in
ija
rb
ar
tv
an
en
do
to
ra
ba
lg
La
Se
m
ol
Uk
Es
Bu
er
Ro
Ar
M
Az
only primary only primary and lower secondary

only upper secondary all
Source: Authors’ analysis based on school-level data obtained from the ministries of education or national statis-
tical agencies of nine ECA countries for various years.
Note: The definition of a “large” school is flexible in order to recognize different types of schools in ECA countries.
A primary school that teaches only grades 1–4 is defined as a large school if it has more than 50 students per
grade and thus a total of more than 200 students. A comprehensive general school that teaches grades 1–12 is
defined as a large school if it has more than 600 students.
transportation for students to reach more distant schools, refurbishing

receiving schools, and providing after-school programs.
Address Overstaffing
School systems in the ECA region have to date focused more on protect-
ing teaching positions than on ensuring fewer, higher paid jobs. In light of
mounting cost pressures, many countries in the region are now trying to
downsize the number of both schools and teachers in their systems. In
addition to increasing class sizes and optimizing existing schools, coun-
tries in the region might explore the scope for retiring over-age teachers.
Specifically, they could use the impending wave of teacher retirements to
implement an orderly reduction in the number of teaching positions,
Box 4.5
Rationalizing Class Sizes in Bulgaria

Recent efforts in the ECA region indicate that significant cost savings can be gained
from rationalizing class sizes. Bulgaria’s recent experience provides the clearest
piece of evidence. Starting January 1, 2007, the Bulgarian government shifted the
financing system for primary education from one based on inputs (i.e., teachers) to
one based on student enrollment. The bulk of financing for education to munici-
palities is now provided in the form of a large earmarked grant based on the num-
ber of students in the municipality. Municipalities can then decide themselves how
many schools they want to maintain. Not surprisingly, many municipalities quickly
decided they could do with one, two, or three fewer schools than they had previ-
ously maintained, together with larger class sizes. As a result, they started approach-
ing the Ministry of Education with requests for school closures.
The Ministry of Education in Bulgaria still plays an important role in the school
closure process. It has the final say in school closings and monitors compliance
with a fairly elaborate set of procedures that municipalities must undertake before
closing a school. For instance, a school closure proposal has to include a discussion
of how the transportation needs of affected students will be addressed and
demonstrate that nearby schools have sufficient capacity to accommodate the
additional students. The government has also introduced a program to monitor
dropout rates at receiving schools. And it has established a number of additional
national programs to support municipalities in the process. For instance, munici-
palities with school closures can apply for school buses, additional payments for
laid-off teachers, and resources to refurbish “central” (or newly “merged”) schools.
Finally, the ministry has put together a list of “protected schools”: schools that
cannot be closed because no nearby schools exist, thus ensuring access to educa-
tion for all. Additional resources to finance such small schools (usually located in
mountainous areas) are also provided by the ministry.
The results of this process in Bulgaria are shown in figures B4.5A and B4.5B be-
low. After many years of unsuccessful attempts to close schools (interrupted by
brief spikes in school closures in 1997 and 2000), Bulgaria has seen a sharp increase
in school closures since per student financing was put in place in 2007. Prior to
these closures, average class sizes had been gradually decreasing even in large
schools. However, with better incentives and greater freedom to take action, the
principals of these schools turned around this trend: the average class size rose for
the first time (albeit only marginally) in 2008–09 (see figure B4.5A).
Box 4.5 (continued)
Figure B4.5A Estimated Number of Public School Closures in Bulgaria from

1991–92 through 2008–09
350 4000
300 3500
number of school closures
3000
250
number of schools
2500
200
2000
150
1500
100
1000
50 500
0 0
2
19 –94
19 –95
19 –96
19 –97
19 –98
19 –99
20 –00
20 01
20 02
20 –03
20 –04
20 05
20 6
20 –07
20 08
9
3
–9
–0
–0
–9
–
–
–
91
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
92
19
19
19
number of closures number of schools

Source: World Bank staff calculations based on Bulgaria National Statistical Institute Web site (www.nsi
.bg/index_en.htm) and estimates for the 2008–09 school year.
Note: The number of “school closures” has been estimated somewhat crudely, as has the difference in the
number of public general education schools (i.e., the “net change” in the number of schools). Thus, in
principle, there could have been more school closures in a given year if new schools were opened that
same year.
Figure B4.5B Average Class Size and Student-Teacher Ratios in Bulgaria,

2000–08
22.4 13.6
13.4
22.2
average (number of students)
13.2
student-teacher ratio
22.0
13.0
21.8 12.8
21.6 12.6
12.4
21.4
12.2
21.2
12.0
21.0 11.8
00
01
02
03
04
05
06
07
80
20
20
20
20
20
20
20
20
20
class size student-teacher ratio
Source: Authors’ calculations based on school-level data obtained from Bulgaria’s National Statistics Insti-
tute Web site, www.nsi.bg/index_en.htm.
while slotting new hires for subjects in greatest demand, such as informa-
tion technology. The use of stronger redundancy or retraining packages or
both for departing teachers might make departures more acceptable. The
experience of countries now in the process of reducing teacher numbers
in response to the ongoing economic crisis (e.g., Latvia and Romania)
may provide valuable lessons in this area.
At the same time, governments in the region will need to develop
strategic policies to reinvigorate the teaching profession (see box 4.6).
Box 4.6
Developing New Teacher Policies

When policy makers in the ECA region are freed from the daily operational work
of managing schools, they can devote more time to managing the most impor-
tant asset in the education sector: teachers. Specifically, they have the opportu-
nity to design long-term policies to restore the prestige of the teaching profes-
sion, attract higher-caliber teachers, and motivate them to remain in the profession.
The importance of good teachers is well understood as a central feature of top-
performing education systems worldwide, which all share one important feature:
the ability to attract high-performing individuals into a teaching career and turn
them into effective instructors.a
One innovation that ECA countries could consider would be to introduce per-
formance as a more formal criterion for teacher remuneration. In Chile, the Ministry
of Education introduced a productivity bonus called the “National Subsidized School
Performance Evaluation System” in 1996. The aim of the program was to improve
academic performance via the bonus, which was given to teachers working in insti-
tutions that showed the best test score results over the course of two years. To
ensure that test score results were compared between similar schools (i.e., schools
with students from similar socioeconomic groups), “homogenous school groups”
were created and the competition between teachers and schools were created
within these groups. The scheme in Chile is thought to have been associated with
generating better learning outcomes (Contreras, Flores, and Lobato 2003).
Another innovation that ECA countries could consider is to more carefully
monitor teaching practices (as opposed to monitoring student test score results
alone). Having standardized information on classroom teaching practices in addi-
tion to test score results would provide policy makers with intermediate informa-
tion on strengths and weaknesses in the learning process.
Box 4.6 (continued)
In the United States, a recent large-scale evaluation of teacher policies relied

on a prominent tool—the Vermont Classroom Observation Tool (VCOT)—that
may offer a suitable starting point for ECA countries. The VCOT reportedly
measures those teacher practices that current research suggests are essential to
good teaching or that have been linked to student achievement. According to
the researchers involved in the evaluation, who reviewed all available tools for this
purpose, VCOT has a number of desirable features, including the following:
• An appropriate level of detail on teaching practices that are deemed to be good

instruction, including the ability to capture complex teacher behaviors, such as
whether a teacher makes connections between reading and writing
• Relatively simple implementation in the field (see Glazerman et al. 2009).
Needless to say, more conventional tools to heighten the quality of future

teachers are also important, such as accrediting teacher training programs and
requiring that teachers be certified before entering the classroom. Best practices
mandate that accreditation be performed by an independent agency and focus
on evaluating the learning outcomes of teacher training programs, rather than
prescribing fixed ways to accomplish prescribed learning objectives (OECD
2005a). Teacher certification, moreover, can be an effective way to assess whether
teaching candidates possess the qualities essential to the profession.
Sources: McKinsey & Company 2007, 2008; Glazerman et al. 2009; OECD 2005a; authors’ analysis.
a. Although perhaps unsuitable for a national strategy, the Teach for America Program in the United
States and the Teach First Initiative in the United Kingdom provide invaluable lessons for policy makers
seeking to develop aggressive teacher recruitment strategies. Both programs employ the same tools
used by top private firms to attract high-caliber candidates (see McKinsey & Company 2008).
Summary
To manage for results at the pre-university level, education ministries
should implement national student assessment exams on an annual
basis, as well as participate regularly in student assessments such as
PISA, TIMSS, and PIRLS. For assessment data to produce the kind of
information needed to improve student outcomes, however, background
information on the students who take these assessments must be col-
lected at the time of the tests. Without such information, the usefulness
of test scores is severely constrained because the data cannot be disag-
gregated by different types of students, such as those from different
socioeconomic backgrounds or ethnic minorities. Policy makers can cre-

ate greater demand for this data in part by making student learning
outcomes a component of teacher and school evaluations.
Perhaps most important, countries in the ECA region need to move
beyond using national assessments solely for examination purposes and
begin using such tests to assess the quality of education at each level and
thus inform policy making in the sector. In addition to performance data,
ECA countries need to regularly collect data on the labor market out-
comes of secondary graduates, primarily through labor market surveys
and periodic tracer surveys of graduates.
Central education ministries should also abandon a “policing” mindset
that manages schools for compliance with norms and instead grant schools
greater autonomy in return for accountability. This means giving lower-
level actors greater decision-making power over school operations and
budgets, as well as relaxing central norms on class sizes. Greater autonomy
is particularly needed in vocational education and training, where pro-
grams need the ability to response to demand. Among the potential
accountability mechanisms that policy makers can implement are the
creation of school councils; school development plans that outline each
school’s strengths and weaknesses; “school scorecards” that include basic
indicators on a school’s performance; and the sanctioning of low-perform-
ing schools and rewarding of high-performing schools, defined—in both
cases—as schools that improve (or not) the learning outcomes of all types
of students, including those from low-income and minority households.
unless policy makers in the region address the neglect of the teaching
force, an effort that will involve resolution of serious overstaffing prob-
lems and improving the pay, professional development, and work condi-
tions of teachers. Specifically, policy makers should take advantage of the
impending wave of teacher retirements to implement an orderly reduc-
tion in the number of teaching positions, while slotting new hires for
subjects in greatest demand, such as information technology. The use of
stronger redundancy or retraining packages or both for departing teachers
may make departures more acceptable.
Finally, per student financing should be introduced in the school sys-
tem using a formula that allows the per student amount to vary according
to the cost of providing education in a given region or school. Such fund-
ing also embeds incentives into financing flows: actors in the system are
rewarded by delivering results and will therefore seek to do so.
Notes
1. For an example of school scorecards, see the Ministry of Education of Chile
website at www.simce.cl. For an example of school report cards with value-
added results, see U.K. school results on the U.K. Department for Education
website at www.education.gov.uk/performancetables./ (both URLs accessed
January 2011).
2. U.S. Department of Education. 1998. “Turning Around Low-Performing
Schools: A Guide for State and Local Leaders.” U.S. Department of Education,
Washington, DC. http://www.ed.gov/pubs/turning/intervene.html. Accessed
September 2010.
CHAPTER 5
Managing for Results in the

Tertiary Education Sector
Despite certain areas of excellence and growing numbers of students, the

quality of tertiary education1 among countries in the ECA region contin-
ues to be of concern. The sector has expanded, but the growth has
occurred without sufficient quality assurance mechanisms and without
the necessary information enabling users to make informed choices. As a
result, it is unclear if tertiary students are graduating with the advanced
competencies needed by future employers.
A number of countries in the region have already granted tertiary pro-
viders considerable autonomy.2 These countries, while continuing to
promote local management of this sector, now face the additional hurdle
of improving academic and fiscal integrity by introducing accountability
mechanisms that focus on improving learning outcomes. Other countries
in the region have yet to change their centralized management practices.
These countries face the simultaneous challenges of enhancing university
autonomy and developing stronger accountability mechanisms. All of the
ECA countries, however, need to make greater efforts to collect data on
student learning and employment outcomes, and introduce performance-
based financing.
Even though fundamental integrity problems remain widespread in
the university sector, it is important that ECA policy makers refrain
139
from attempting to resolve these problems before they devolve authority

to universities and introduce flexible financing. The great temptation in
the region—given its tradition of strong centralized control—is to fix
integrity problems with even stronger centralized control and manage-
ment, which history suggests is unlikely to improve the situation. Even
if it could, more control is unsuited for improving either relevance or
financial efficiency in the tertiary sector. Instead, greater autonomy is
needed if local decision makers are to discover and adopt local, innova-
tive solutions that equip their graduates with the competencies needed
on local labor markets.
In addition to strengthening accountability, all ECA countries need to
focus on collecting more student outcome data. Here, the region shares
the problem of advanced industrial nations: an inability to evaluate sector
performance based on student learning outcomes. Performance in tertiary
education continues to be primarily evaluated based on the data that are
easy to generate, namely, the number of academic publications by faculty,
which says very little about the quality of teaching and even less about
how much students have learned. The only way to shift the discussion of
performance towards student learning outcomes is to start using stan-
dardized tests to measure what students learn.
Greater use of performance-based financing in the sector is also critical
for improving the accountability and efficiency of higher education insti-
tutions. In addition, more private resources will be needed to help the
tertiary sector meet rising student demand, with fee-based mechanisms
offering an additional way to strengthen the focus on results.
Introduce Learning Assessments and Track

Employment Outcomes
As in the pre-university sector, the need for better outcome data in the
tertiary sector is the first priority of central governments in the ECA
region because without such data, they cannot address the legacy of cen-
tral planning. That is, they cannot move from micromanaging inputs to
steering the system through policy guidance and standards. Specifically,
policy makers need to know whether scarce public resources (and exist-
ing rules and regulations) in the university sector are working or need
adjustment. For such an assessment, a range of different data is required,
with data on student learning and employment outcomes the most
important. Only this type of data can answer the question: What is the
sector achieving in terms of results?
Managing for Results in the Tertiary Education Sector 141
Governments are not the only actors that would benefit from better
information on outcomes. Employers, students, and parents also need
objective, standardized data to make informed decisions. The market for
tertiary education is a classic example of the seller (i.e., universities) being
vastly better informed than their consumers (i.e., students, parents, and
future employers), or even their financiers and donors (i.e., governments,
for the most part) (Carey 2010). When students make choices about
future universities and degrees, they usually do so without having met
the professors who will teach them and with little to no understanding
of the skills that they will acquire. This asymmetry places a great deal
of power in the hands of the seller, which can contribute to worsening
quality and higher costs for the consumer. In other parts of an economy
characterized by similar informational asymmetries, policy makers usu-
ally play a strong regulatory role or mandate that providers provide
better information to consumers or both.
Currently, data on tertiary outcomes in the ECA region as elsewhere
focus on research outcomes, such as how many publications the faculty of
a university is producing. Unfortunately, when this is the only performance
indicator available, it inadvertently receives more attention than it should.
Indeed, rankings of universities in the region place significant weight on
the number of publications that they produce. However, from a skills per-
spective, the outcome that matters most is whether students are graduat-
ing with the competencies in demand on the job market and finding jobs.
Currently, these outcomes are not being measured. Governments in the
ECA region can change this situation by granting tertiary institutions pub-
lic resources in return for them providing useful information about stu-
dent learning and employment outcomes. There are several complementary
ways of gathering this information, some of which are more direct than
others. These include rankings or league tables, tracer studies, and direct
measurement of student competencies via standardized tests.
As noted in earlier chapters, standardized tests are the best available
indicator for measuring student performance, especially when socioeco-
nomic characteristics are collected for each student (allowing the results
of different socioeconomic groups to be compared). Such data are espe-
cially useful if organized to track the progress of individual students over
time. Together with other indicators, these tests provide data on differ-
ences in learning outcomes between weaker and stronger students and
institutions. Without this information, policy makers have no way of
identifying—and thereby rewarding—institutions that do the best job of
giving the largest number of students a quality education. Unfortunately,
Box 5.1
The Difficulty of Measuring Competencies at the

Tertiary Level
It is perhaps understandable that higher education institutions around the world
have been reluctant to let governments interfere in designing tests and measur-
ing competencies of their students, arguing that such attempts would interfere
with academic freedom. Any attempt to introduce such tests should accordingly
be carried out in cooperation with higher education institutions.
Several countries around the world have introduced standardized tests to mea-
sure what competencies their tertiary students acquire during the course of their
studies. For instance, in the United States, the Collegiate Learning Assessment (CLA)
was developed in the 2000s by a subsidiary of the RAND Corporation, and is cur-
rently used only by roughly 400 institutions. Recognizing that students choose very
different academic specialties in college, “the CLA tests the higher-order thinking
skills that all college graduates should possess: critical thinking, analytic reasoning,
and communication. The exam is given to a sample of freshmen and seniors to esti-
mate how much students learn in college” (Carey 2010, 16–17).
Similarly, since 1949, the Graduate Record Examination (GRE) has been used as
an admission requirement by numerous graduate programs in the United States.
It is a standardized test that seeks to measure verbal and quantitative reasoning,
analytical writing, and critical thinking skills.
The OECD is currently preparing an assessment of tertiary students called the
Assessment of Higher Education Learning Outcomes (AHELO), a tool that will “as-
sess learning outcomes on an international scale by creating measures that would
be valid for all cultures and languages.”a However, the OECD does not expect a
full-scale AHELO to be launched before 2016.
Sources: Carey 2010; author’s analysis.
Note: a. See the AHELO website for more details: www.oecd.org/edu/ahelo (accessed September 2010).
a standardized international learning assessment for tertiary students is not

expected to be launched until 2016 (see box 5.1). Thus, to gather the
necessary data now—when it is needed—ECA countries will need to
begin with their own, domestically developed approaches. Several coun-
tries around the world have already started introducing standardized tests
to measure student learning outcomes (see Salmi 2011 for a review). Such
efforts would provide complementary information to the international
comparisons that Assessment of Higher Education Learning Outcomes
(AHELO) will generate in the coming years.
Experiment with Measuring Learning Outcomes

All countries in the ECA region save Tajikistan, the Kyrgyz Republic,
Uzbekistan, Turkmenistan, and Kosovo, have become signatories to the
Bologna Process and the push to create a European Higher Education
Area (EHEA) by 2010.3 As participants in the Bologna Process, most
ECA countries have taken important steps towards establishing and
strengthening quality assurance institutions, and, most important, mov-
ing towards an environment where university degrees are described
based on learning outcomes, competencies, and student workloads
(Adelman 2003). “Learning outcomes” in the Bologna process are generic
statements of what a learner knows and is able to do at the end of a
period of study. This initiative has established learning outcomes for the
three levels, or cycles, of tertiary education (bachelor, master, and doctor-
ate), as well as for so-called “short degree” or “short higher education”
programs.4 Important preparatory work in this area has been done by the
Tuning Project,5 which brought together academics from all over Europe
to develop subject-specific learning outcomes for a range of professions,
including, for example, history and nursing.
The establishment of a European Qualifications Framework for Lifelong
Learning, followed by the national qualifications frameworks (NQFs) of
individual countries, are other important milestones toward the measure-
ment of learning outcomes. Within EHEA today, qualifications frameworks
and the European Credit Transfer and Accumulation System (ECTS)
allow teachers and students to establish how many “credits”—the quantita-
tive indicator for units of learning—a student obtains from finishing a
course, program, or other unit of learning. However, countries in the
EHEA have not yet moved toward introducing standardized assessments
to measure what students acquiring these credits can do in terms of skills.
One way that ECA countries might begin to measure the learning
outcomes of tertiary students would be to introduce standardized testing
within fields of study where these outcomes have been clearly defined.
Alternatively, ECA ministries of education might develop tests of the
broader competencies that all tertiary graduates can be expected to have
(e.g. verbal reasoning, quantitative abilities, analytical writing, and criti-
cal thinking skills). The OECD’s DeSeCo project (OECD 2005b), dis-
cussed in chapter 2, could provide a starting point for the definition of
such broad competencies.
Start with rankings, tracer studies, and student surveys. Although it is

likely to take years before the competencies of tertiary students can be
measured using standardized tests, policy makers in the ECA region can
take immediate steps to provide more information about the quality of

the tertiary sector. First, they can mandate that all tertiary institutions
survey their students after graduation, using graduate tracer studies. Such
studies can be implemented relatively quickly and would provide initial
insight into a system’s current strengths and weaknesses. Second, regular
surveys on the practice of purchasing admission and cheating may also be
worth considering, given the magnitude of reported fraud in the tertiary
sector (see the next section), even though such surveys would not be a
measure of quality per se. Third, regular (standardized) surveys of stu-
dents regarding their satisfaction with their university choice and the
teaching at that institution could also be implemented relatively quickly
and provide additional information for the compilation of rankings.
With respect to tracer studies, two countries in the region—Hungary
and Romania—are making rapid progress. In fact, these studies are
becoming a core element of tertiary management in these countries. In
Hungary, 2010 marked the first year that results from the new “Graduate
Career Tracking System” were produced (spanning graduates from 25–30
institutions). This project is the culmination of work that began in 2008
and involved 30 ongoing EU-financed projects. In Romania, data from a
graduate survey will be available in 2011. By that time, policy makers will
have results from surveys of students who graduated in 2008–09 (i.e., 12
months after they graduated), as well of students who graduated in
2004–05 (i.e., five years after they graduated). As is the case with
Hungary, the development of the tracer study in Romania is being
financed by EU Social Funds.
Hungary is moving ahead rapidly on tracer surveys for several reasons.
In the first place, central policy makers in the country are pressuring
tertiary institutions to start collecting such data. The 2005 Higher
Education Act of Hungary, for example, makes it mandatory for every
university and college to carry out surveys of graduates. The central gov-
ernment has also made tracer studies a part of quality assurance discus-
sions, with the availability of such surveys now (or soon to be) tied to
institutions’ accreditation agreements. In addition, the government is
using the power of the purse, tying tracer survey data (or their availabil-
ity) to three-year financing agreements. In the second place, institutions
of higher education in Hungary themselves consider tracer data useful for
a number of reasons: (1) they want labor market feedback to help them
design better programs; (2) the data can be used in marketing; (3) the
data is valuable for internal quality assurance; and (4) tracer surveys are
one of many engagement tools for strengthening an alumni network.
Rankings have also played an important role in helping broaden what

performance in the higher education sector means, moving the definition
well beyond a simple measure of the total number of publications a uni-
versity faculty produces. For instance, the British newspaper The
Guardian provides “University League Tables” on 46 different fields of
study (from programs focusing on “agriculture, forestry, and food” to
“veterinary science”) that includes measures of student satisfaction with
courses and teaching.6
Conceptually, rankings are fundamentally a less valid performance
indicator than student assessments or surveys of employment outcomes
because they focus primarily on inputs and processes, rather than student
outcomes. Nevertheless, rankings play a powerful role by providing cer-
tain key information on institutions, strengthening competition among
them, and highlighting the idea that “performance” matters.7 For example,
the multidimensional ranking of the Center for Higher Education
Development (CHE) (see Usher and Medow 2009), has gained a strong
international reputation. The risk, however, is these efforts promote com-
petition along the dimensions—that is, inputs and processes—that may
not necessarily be relevant to outcomes, thereby possibly diverting atten-
tion and resources away from the outcomes that really do matter.
Introduce greater autonomy and encourage private sector participation.

As noted at the outset of this chapter, authority for operational manage-
ment of tertiary institutions has been most fully devolved in the new EU
member states and to a lesser degree, in the countries of South Eastern
Europe and in Ukraine and the Russian Federation (see box 5.2 for a
description of this process in Romania). By contrast, the reform process
has barely begun in Belarus and many other CIS countries. In countries
where more freedoms have been granted, institutions of higher education
have begun to act in a more entrepreneurial fashion, align their programs
to a greater degree with the demands of students and the labor market,
and respond efficiently to government incentives.
University autonomy is not, however, generally expanded through a
straightforward transfer of authority between two static entities: the
national government and tertiary institutions. Rather, tertiary institutions
themselves have changed as governments have developed new methods
of providing guidelines—a trend that has occurred more to date in
Europe and North America than in the ECA region. In his recent study,
Usher (2009) notes that governments have not simply handed authority
over to traditional self-governing universities, but rather, transferred
Box 5.2
Decentralization of the University Sector in Romania

The degree of autonomy granted to tertiary institutions in Romania rapidly
expanded between 1995 and 2005, a period that saw less state regulation, more
academic self-governance, and greater managerial governance at the university
level. Stensaker, de Boer, and Enders (2006) have recorded the changes in these
three criteria for a number of ECA and non-ECA countries. When judged against
these standards, Romania’s tertiary institutions (along with those of many other
ECA countries) have experienced a large increase in autonomy. The financing of
higher education in Romania also underwent major reform, shifting from input-
based funding (i.e., linked to the number of professors employed) to student-
based funding (i.e., linked to the number of students enrolled).a
Increases in autonomy and flexibility in financing happened very quickly—
before nascent measures aimed at holding institutions accountable could take
hold. For instance, the rapid expansion in enrollment that began in 1990 occurred
when the newly established Romanian Council for Accreditation (CNEAA, estab-
lished in 1993), was still discovering its mandate and trying to establish its institu-
tional credibility. Many new (mostly private) institutions were established during
the boom years, but CNEAA simply did not have the capacity to review and
accredit every new program or institution; that task was, in some instances, out-
sourced to public universities, which—in exchange for a fee—were charged with
taking a private, growing institution under their wings. By 2005, the shortcomings
of CNEAA had become apparent and it was replaced with a new quality assurance
agency—the Romanian Agency for Quality Assurance in Higher Education
(ARACIS)—modeled on the European Standards and Guidelines (ESG) for quality
assurance (see Korka 2008).b
There is visible evidence that the accountability framework in Romania today
remains insufficient. For example, a long-lasting tug-of-war continues between
the Ministry of Education, Research, and Innovation (MERI) and the largest private
university (Spiru Haret University)—where as many as one-eighth of all university
students are enrolled—over the right to enroll students without a license.
Although MERI has refused to accredit certain long-distance learning programs
because of a lack of adequate professors, the private university continues to enroll
and graduate students from these programs. There are also widespread integrity
problems in the sector, as revealed by the Coalition for Clean Universities in a 2009
report, “University Integrity Contest.” As discussed in more detail in box 5.4, the
Coalition identified three concrete problems that undermine universities’ ability

Box 5.2 (continued)
to deliver quality education: increased tolerance of plagiarism, extended nepo-

tism, and a lack of transparency in decision making and the academic process.
In response to these weaknesses, the government introduced a new law on
education in 2011. The new law imposes many restrictions on the administrative,
financial, and staffing autonomy of universities. At the same time, the new law
includes key measures to strengthen accountability for performance. For instance,
the law requires extensive reporting by the universities and much transparency in
their operations. Importantly, the law also introduces new ways to sanction public
universities (including by withholding public funds).
Sources: Stensaker, de Boer, and Enders 2006; Korka 2008; Nastasescu 2006; Coalition for Clean
Universities 2009.
Notes: a. As noted earlier, this book refers to a student enrolled as an “output” (not an “input”) and teach-
ers, textbooks, books, financing, and so forth, as “inputs.”
b. One of the differences between CNEEA and ARACIS was that CNEEA was a consultative body of
the Ministry of Education, while ARACIS was established as an independent body. Another difference
between the two bodies is that CNEEA did not involve international peers, whereas ARACIS does
(see Nastasescu 2006).
authority to an entirely new managerial level that largely superseded

former governing entities. At the same time, these governments have put
greater emphasis on quality assurance and accountability structures at the
university level. According to some authors (see Santiago and Tremblay
2008), this shift in emphasis enhances, rather than reduces, their ability
to direct higher education systems. Thus, while they have increasingly
devolved day-to-day decision-making powers to a new level of university
manager, governments have intensified their oversight of higher educa-
tion. This is now the central task before the group of ECA countries that
have already largely decentralized their university systems.
Low-income CIS Countries Need to Initiate Reform

In ECA countries that have maintained tight centralized control over the
tertiary sector, the task ahead will be to start devolving authority to insti-
tutions of higher education. In some of these countries, however, the very
opposite has occurred, with governments now forcing private institutions
to comply with a host of new curricular and budgeting regulations.
Ministries would instead benefit from delegating detailed operational
responsibilities (e.g., controlling the number of student seats in various
programs, managing budgets) and focusing on monitoring educational
quality, mobilizing resources, and designing policy to improve strengths
and reduce weaknesses in the tertiary education.
Azerbaijan is a good example of a country in which the tertiary sec-

tor remains highly centralized (see Salmi 2009). Very little autonomy
has been granted to public universities in the country and there has
been virtually no growth in the number of private universities. The
Ministry of Education continues to centrally control student intake at
every university in the country, even in the few private ones that exist.
The ministry also decides which programs a university may open and
enforces the closing of programs in areas perceived to be either
saturated or of little relevance. For example, in 2006, a number of uni-
versities had to terminate their programs in law, medicine, and interna-
tional relations.
The financing of higher education in Azerbaijan has also not been
reformed and continues to be based on historical norms, with funding
tied to the number of professors. In terms of strengthening quality assur-
ance (an important component of strengthening accountability for
results), the Bologna scorecard ranks Azerbaijan’s progress as very lim-
ited, with a score at the very bottom of the list of 22 ECA countries that
participate in the Bologna Process.8 Within this restrictive framework, it
is very difficult for more dynamic tertiary education institutions to
emerge and expand.
It is important that this group of ECA countries not take the path of
the group that provided autonomy first without accountability, but
instead introduce the two as mutually reinforcing policy instruments.
This is not only considerably more effective, it is also politically more
feasible to implement, as it is easier to grant the two together than to
try to impose accountability later on already autonomous institutions.
In carrying out this reform, policy makers in these countries need not
simply transfer autonomy to static, traditional universities but can
instead mandate that increased autonomy be accompanied by more
“businesslike” and accountable leadership and management. Such
change should aim, among other things, at professionalizing institu-
tional governance and management (Stensaker, Enders, and de Boer
2006). A number of countries (or autonomous regions within coun-
tries) provide examples of how this can be done. For example, in
Denmark and Norway and in Quebec (Canada), the wider tertiary edu-
cation community is held accountable by university boards that have a
majority of outside members and the power to hire and fire the leaders
of individual institutions (Fielden 2008, as cited in Salmi 2009). Recent
reforms in Lithuania are moving the governance of higher education in
this direction (see box 5.3).
Box 5.3
Introducing Businesslike Leadership and Management in

Lithuanian Universities
Policy makers and university managers in Lithuania know that it is not easy to get
the right balance between institutional autonomy and accountability. Lithuania’s
Law on Higher Education of 2000 was amended six times during the period
2000–09. On April 30, 2009, additional sweeping reforms were introduced when
an entirely new law on science and higher education was adopted.
The 2009 reforms aim to tackle a common problem of countries that have
granted more institutional autonomy to universities: How do you get more busi-
nesslike leadership and management from university leaders and make them
more accountable to outsiders? Prior to the 2009 reforms, the crux of the problem
in Lithuania—as in many other countries of the ECA region—was the position of
rectors. Rectors were usually elected from within the university; consequently,
they felt accountable only to other faculty members (whose ranks they expected
to rejoin after having served their term). They also usually had limited managerial
experience. Reformers in Lithuania recognized that changing the way rectors
were elected (and to whom they were accountable) was not a “magic bullet” for
getting better university management, but without this change, major reform of
university governance was unlikely.
The essence of the 2009 reforms was twofold. First, it narrowed the mandate
of the Senate by clarifying that its main task was to safeguard and promote the
academic integrity and prestige of the university (i.e., approving study programs
and ensuring admittance follows academic standards), not to manage it. Second,
it made explicit that rectors were accountable to a broader set of stakeholders
than just faculty members, one that included taxpayers. This was achieved by
changing the way rectors are nominated and selected. In the future, the Govern-
ing Council—two-thirds of whose members are from outside of the university—
will establish an election committee and hold open competitions to search for
suitable candidates. (See table B5.3 for a summary of the reforms.)
In exchange for these reforms, university leaders have sought even greater
institutional autonomy from the central government. In particular, they pointed
to inherent inconsistencies in the major pieces of legislation governing higher
education. For instance, although Article 5 of the Law on Higher Education states
that the activity of higher education establishments is based on academic free-
dom and autonomy, the same article also says that this activity is determined
Box 5.3 (continued)
Table B5.3 Summary of Changes

Governing body Before 2009 reforms After 2009 reforms
Governing Council Role: Role:
Supervisory; confirms Highest governing body
that Senate and Rector responsible for strategic
conform to the Statute decisions
and whether the activity Composition of Council:
and administration of the Same composition, but with
university are transparent greater efforts to select more
Composition of Council: business leaders as Council
2/3 appointed by Ministry members
of Education New mandate:
1/3 University representatives Establish an election
(including Rector and committee and have open
student leaders) competition for selecting
Rector
Rector Role: Role:
Highest administrative office Unchanged
responsible for the
management of the
university
Senate Role: Role:
Highest self-governance Highest body dealing with
institution in the university, academic matters; approves
responsible for its general study programs and ensures
affairs (including selection admittance follows
of Rector) academic standards
Elect Rector
Source: E-mail correspondence of authors with Ministry of Education of Lithuania, 2009.
not only by this law and the statute of the institution, but also by other laws. In
practice, this meant that universities could not own their buildings or borrow
funds, and were only able to spend budgets as they saw fit to a limited extent. As
part of the 2009 reforms, these inconsistencies are being addressed by changing
the legal status of universities from a budgetary to a public entity. This will provide
them more freedom and the right to own property, as well as expand their rights
to manage the property entrusted to them by the state.
Create an Accommodating Environment for Private Providers

Another possible way to create a more flexible university sector is to
provide an accommodating environment for private providers to establish
universities and flourish. Although clearly not a panacea—and not with-

out some attendant integrity problems (see below)—private resources are
nevertheless critical because the cost pressures facing higher education in
the region are simply too great. Private providers can be advantageous
because they are usually (although not always) nimble and responsive,
being less constrained by a top-heavy bureaucracy and political factors.
They are also often run by especially motivated and entrepreneurial indi-
viduals who, in many instances, establish an institution because they see
unmet demand. This type of creativity can prove very valuable to an
education and training system that is seeking to become more responsive
to rapidly changing labor markets—provided, of course, that adequate
quality standards are enforced. (For more on the subject of private pro-
viders and private funding for tertiary education, see the section on
financing below.)
Private providers in many ECA countries already play a crucial role in
absorbing the increased demand for tertiary education that public provid-
ers have been unable to meet. Indeed, the rapidly growing and vigorous
role of private providers in the ECA tertiary sector is one of
the more notable features of the region’s educational systems. While the
private sector at the tertiary level is not equally strong across the entire
region (in a number of countries, particularly Croatia and the Czech
Republic, private education is notable by its almost complete absence),
it generally has a much stronger presence than in Western Europe,
accounting for as much as one-third of total enrollment in some
countries. For example, in Romania, private universities have played a
crucial role in accommodating a tripling of the number of tertiary stu-
dents since 1998. By 2007–08, private enrollment there accounted for
approximately 40 percent of all students enrolled at this educational level
(World Bank 2008b).
In fact, the boom in private tertiary education in the ECA region is a
characteristic example of governments using the nongovernmental sec-
tor to absorb demand (Slantcheva and Levy 2007). Most countries that
embraced this solution have created a large number of small institutions
that specialize in subjects that are inexpensive to teach (primarily law,
social sciences, economics, and business administration) and whose
teaching staff consists primarily of academics from the public sector
who teach part-time to supplement their incomes (see, for instance,
Linden, Arnhold, and Vasiliev 2008). This has been an efficient way
to meet the niche demands of specific employers, satisfy demand
in smaller and more remote areas of countries where public institu-
tions are not available, and introduce new (usually foreign) teaching
techniques and curricula to a local market (for example, New York

University in Tirana).
These private institutions have, to a certain degree, even taken part in
attempts to change the prevailing culture with respect to transparency,
markets, and democracy (for example, Khazar University in Azerbaijan
and the European Humanities University in Belarus). However, private
sector providers have not yet been able to offer education in the sciences
or engineering or to provide education in a research setting. Throughout
the ECA region, these specializations have remained the preserve of the
public sector, with the notable exception of the British-Kazakh University
in Almaty.
Strengthen Accountability
Accountability at the most basic level refers to fundamental academic
and fiscal integrity, which touches on such issues as admissions and finan-
cial fraud, plagiarism, and professorial nepotism. Here, almost all ECA
countries continue to have serious problems. Strengthening accountabil-
ity in tertiary education also entails ensuring that public resources are
spent effectively, the education provided is of high quality, and that study
programs are relevant to students’ future in the workforce. Although it
might seem necessary to address basic integrity issues before dealing with
higher-order accountability issues, all of these problems can be addressed
concurrently, a task made easier by the fact that most basic accountability
instruments also address higher-level concerns.
Improve Fiscal and Academic Integrity

At the lowest level, strengthening accountability in higher education
means safeguarding the system’s basic integrity. With regard to finances,
this implies preventing embezzlement, fraud in public tenders, collusion,
and so forth. With regard to academics, it means avoiding examination
fraud, unethical behavior among faculty, noncompliance with admis-
sion standards, research fraud and other forms of plagiarism, and decep-
tion in the quality assurance process (Salmi 2009).
With the exception of a few new EU member states, evidence suggests
that higher education across the region is still struggling with basic integ-
rity issues, both academic and financial. As discussed in chapter 2, this
evidence comes in different forms. For example, when randomly selected
university students in a number of ECA countries were asked a range of
different questions related to purchasing grades, admission, and diplomas,
more than 60 percent of respondents reported knowing of other students

who had purchased either entrance to the university or a specific grade
(see figure 2.6). Newspapers and other media regularly report on fraud
and corruption in the sector. For instance, a scandal erupted in the Czech
Republic in 2009 following revelations that a number of students had
been awarded law degrees by the University of West Bohemia in Pilsen
after only a few months of study. The scandal led the Minister of
Education to order a national audit of all university degrees awarded since
2000, covering some 315,000 graduates (see Holdsworth 2009).
The fact that such problems persist is also evident in the work of
national quality assurance and accreditation agencies. For instance, in
Georgia, when a National Education Accreditation Center was finally
established in October 2004—13 years after the first private provider
was allowed to operate—its first assessment of the 178 existing institu-
tions of higher education (both public and private) brought sobering
news: only 78 passed the minimum quality requirement established by
the center. While both private and public institutions were barred from
admitting students in 2005, the vast majority of barred institutions
were private. During the years when there was little or no regulation
of quality, an estimated 20–30 percent of all tertiary students gradu-
ated from unregulated—and largely unmonitored—private institutions
(Pachuashvili 2007).
As a recent report (Salmi 2009) documents, countries around the
world utilize a range of different instruments to address basic integrity
problems and strengthen accountability for educational quality and
fiscal efficiency. The good news is that more sophisticated instruments
generally help with accountability problems across the board, that is,
with both basic integrity and higher-order accountability concerns.
The various instruments available and their uses are summarized in
table 5.1.
Many new members of the EU in the ECA region are already familiar
with the instruments used to strengthen accountability for better learning
outcomes. Yet these tools also strengthen basic integrity. One such instru-
ment is licensing, which is critical for strengthening academic integrity
in tertiary institutions. Other tools that promote both academic integrity
and the quality of education include regular institutional evaluations,
such as accreditation procedures, academic audits, and other forms of
evaluation. Public disclosure laws are also useful. In addition, it is possible
to foster programs in which outsiders—or anyone with no vested interest
in the university system—review an institution’s academic integrity (this
Table 5.1 Tools for Strengthening Basic Academic and Fiscal Integrity in
University-Level Institutions
Tools that strengthen academic integrity Tools that strengthen fiscal integrity
• Licensing • Financial audits
. . . and educational quality . . .and fiscal efficiency
• Accreditation, academic audits, and • Public disclosure laws
evaluations
• Public disclosure laws • Strategic budget plans
• Fostering of outsiders to review academic • Performance contracts based on
integrity (e.g., Romania) performance indicators
• Performance contracts based on • Student loans, scholarships, and vouchers
performance indicators
• Creation and dissemination of rankings, • Embedding of incentives into allocated
benchmarks resources (e.g., tying funding to the
number of students enrolled or graduated,
rewarding good performers, etc.)
Sources: Authors’ review of available instruments, as presented in Salmi 2009.
tool been used in Romania, see box 5.4). Performance contracts based
on performance indicators can also help ensure integrity and educational
quality. Finally, it can be valuable to facilitate the creation of rankings or
benchmarks or both and then disseminate these data in an appropriate
manner.
Instruments that address basic fiscal integrity also have an impact on
the efficient use of resources. One obvious tool for improving basic finan-
cial integrity is the financial audit. Tools that accomplish both fiscal integ-
rity and the efficient use of resources include public disclosure laws,
strategic budget plans, performance contracts, external watchdogs (see
box 5.4), and student loans, scholarships, and vouchers. In addition, it can
be useful to embed incentives into the resources allocated to higher edu-
cation institutions, for example, by linking funding to the number of
students enrolled or graduating or using it to reward good performers.
These suggestions are discussed in greater detail in the section on financ-
ing below.
Implement Immediate and Long-term Quality

Assurance Mechanisms
Strengthening accountability for educational quality is an especially long
and difficult process. At the outset, setting up effective accreditation
or quality assurance mechanisms requires building a country’s capacity
for these tasks. In view of the relatively weak institutional capacity for
Box 5.4
Using External Watchdogs to Shed Light on

Integrity Problems
In 2007 in Romania, a group of 14 nongovernmental organizations (NGOs)
combined forces to form the Coalition for Clean Universities (CUC) to monitor
Romanian public universities and establish norms for good governance. The
CUC piloted a methodology to assess the academic integrity of these universi-
ties. A questionnaire was designed and teams of external evaluators—com-
posed equally of experts and students—set about requesting information from
the universities (see Mungiu-Pippidi 2009 for a detailed description of the
methodology). In its widely disseminated report, the CUC emphasized three
findings:
Evidence of increased tolerance for plagiarism. Despite numerous scandals

regarding plagiarism, universities have no tools to control this phenomenon. With
the exception of a single university, evaluators could not identify procedures for
combating plagiarism, either at the diploma level or the level of papers elaborat-
ed by research staff.
Evidence of extended nepotism. In a number of universities, evaluators

identified the existence of so-called “academic families,” raising serious
questions about the objectivity of promotions and evaluations among col-
leagues. In one university evaluated by the CUC, eight pairs of academic fami-
lies existed: three husband-wife pairs and five father-son pairs. Taking into
account the total number of teachers (45), the incidence of “academic families”
was very high.
Lack of transparency in decision making and the academic process. CUC

found that the competition for certain teaching or academic positions is kept
quasi-secret; in the most common cases, there was, in fact, a single candidate.
Similarly, the procedure for approving a university budget (revenues and
expenditures) is often carried out without any real consultation with the univer-
sity’s senate. Promotions and pay increases are also governed by nontranspar-
ent procedures, without clear benchmarks. Evaluators have also witnessed
problems in the publishing of asset and interest declarations, as well as the
absence of declarations on cooperation with the former secret police.
Restrictions in accessing public records of public acquisitions are also a major
problem.
Box 5.4 (continued)
The lack of transparency regarding internal procedures and administrative or

academic results or both can be discovered simply by trying to access university
websites, as they frequently do not include such information on such topics as
employment opportunities, teaching jobs, teacher performance, program con-
tent, or decisions of internal governing bodies.
Source: Coalition for Clean Universities 2009.
monitoring educational quality and relevance among universities and

ministries of education in the ECA region—including newly established
or nonexistent quality assurance agencies—policy makers need to employ
many different, complementary tools to strengthen accountability for
results (see box 5.5). These tools include rankings and tracer studies
(which also provide useful data on institutional performance in the
absence of standardized assessments), together with an enabling environ-
ment for private sector participation. Counting on a quality assurance
agency to quickly establish the capacity needed to externally monitor a
rapidly growing sector is risky—this kind of agency needs to be part of
the solution, but not the only response.
In 2003, the ministers responsible for the Bologna Process began a
consultation designed to lead to a common, but not unified, system of
quality assurance in the tertiary education sector. Two years of consul-
tations between quality assurance agencies, higher education institu-
tions, and student representatives followed, resulting in the adoption in
2005 of the European Standards and Guidelines for Quality Assurance
(ESG) in the European EHEA.9 The main features of this consensus
are the following:
• Tertiary institutions have primary responsibility for quality and are re-
quired to have processes of internal quality assurance.
• Tertiary institutions are subject to external oversight by an agency
charged by the government with assuring the competency of quality
assurance mechanisms.
• Quality review agencies themselves submit (on a voluntary basis) to
quality assessment procedures through the European Quality Assur-
ance Register, which is a joint project of the European Network of
Quality Assurance Agencies (ENQA), together with the European
Box 5.5
Two Approaches to Strengthening Accountability

As noted by Alex Usher (2009), two approaches have been used throughout the
world in the past 20 years to improve tertiary educational quality by strengthen-
ing accountability. The first approach is generally referred to as “quality assurance”
or “accreditation,” and has traditionally focused on ensuring that certain minimum
levels of resources (i.e., inputs) and standards are present to ensure a “quality” edu-
cation. In addition, the approach seeks to put requirements in place so that ter-
tiary institutions themselves monitor the quality of their education.
The second approach focuses on learning conditions and learning outcomes.
This method includes rankings systems and systems of performance indicators. It
also relies to a much larger extent on using students and parents to exert outside
pressure on tertiary institutions to deliver results. Thus, parents and students,
rather than a state agency, become the driving force behind better
accountability.a
These two approaches to enhancing accountability are complementary to,
and not substitutes for, one another. As Jongbloed (2008) has noted, quality
assurance is the equivalent of a restaurant health inspector, while university
rankings are the equivalent of a Michelin guide. Both have their place and nei-
ther can replace the other.
Sources: Usher 2009; Jongbloed 2008.
a. In the terminology of the World Development Report 2004, the first approach is the “long” route to ac-
countability, while the second is the “short” route (World Bank 2003c).
University Association (EUA), the European Association of Institutions

in Higher Education (EURASHE), and the European Students Union
(ESU).
In principle, this structure means that national quality assurance bod-

ies now coordinate to determine mutually acceptable evaluation frame-
works, and thus, visions of institutional quality. At the same time,
institutions are empowered to assess themselves within the framework of
wider agreements on institutional quality and evaluations that their rep-
resentative organizations have worked to develop. This situation is
indicative of a broader governance trend (and changing government role):
increased institutional operational autonomy coupled with strengthened
webs of external coordination.
Regular stock-taking exercises are carried out to assess whether

Bologna participants are meeting concrete measurable goals towards the
bigger goal of creating an EHEA. That is, whether they are making aca-
demic degree standards and quality assurance standards more comparable
and compatible throughout Europe (see Rauhvargers, Deane, and Pauwels
2009). The last such stock-taking exercise took place in 2009 and revealed
that many ECA countries have a considerable way to go in terms of
strengthening quality assurance mechanisms. In particular, the exercise
showed that many countries in the region have not yet incorporated
international peer reviews—one of the three key benchmarks in the
stock-taking exercise—into their quality assurance processes. Figure 5.1
groups ECA countries into three categories, according to their Bologna
scorecard for implementing quality assurance mechanisms.
Figure 5.1 Progress on Implementing Quality-Assurance Mechanisms:

The Bologna Scorecard
fully functional
system of QA with
peer review of QA
functional system
agency, and
of QA covering all
significant student
institutions and with
and international
beginning of a substantial student
participation in
national QA and international
QA process
system with some participation
student and in QA
international
not started or in participation in
early stages of Croatia
QA process
development of Czech Republic
quality assurance Albania Hungary
Armenia Latvia
Bulgaria Lithuania
Estonia Poland
Georgia Romania
Azerbaijan Moldova Macedonia, FYR
Bosnia and Russian Federation Turkey
Herzegovina Serbia
Montenegro Slovenia
Slovak Republic Ukraine

Note: QA = quality assurance.
The 2009 stock-taking exercise also revealed significant weaknesses in

establishing internal quality assurance mechanisms.10 For example, some
countries (both ECA and non-ECA) treat internal quality assurance
within institutions of higher learning only nominally (i.e., having institu-
tions prepare a “self-assessment report”), thereby ignoring an essential
part of the ESG that focuses on “learning outcomes-based and improve-
ment-oriented internal quality assurance systems” (Rauhvargers, Deane,
and Pauwels 2009, 51). A key failing of many countries is that they have
still not managed to induce higher education institutions to describe pro-
grams in terms of learning outcomes. Even fewer countries have moved
to introduce “student assessment procedures designed to measure
achievement of the intended learning outcomes” (Rauhvargers, Deane,
and Pauwels 2009, 55).
Introduce Performance-Based Financing and

Encourage Private Funding Resources
Financing can be a central government’s most potent policy instrument
for steering education providers. First, budget discussions—whether
annual or pluri-annual—are recurrent events that give policy makers a
regular opportunity to reward or discourage managerial behavior.
Granting more autonomy or strengthening accountability instruments, on
the other hand, are more protracted and lengthy processes. Second, in
most ECA countries, public resources continue to provide a large share
of tertiary education funding, giving the central government an important
seat at the table. Finally, public resources can be used to influence the
behavior of private institutions that may not otherwise have formal
reporting arrangements with the government. In turn, private resources
can strengthen a results-oriented outcome by encouraging greater com-
petition in the sector and increasing students’ stake in the end result of
their educations.
Per Student Financing

Per student financing is the simplest and most effective instrument for
moving toward output- and outcome-based financing in ECA countries.
Per student financing, combined with autonomous management of both
public and private education institutions, creates a competitive environ-
ment in which students select the best institutions that most closely meet
their educational goals. Creating built-in incentives for individual institu-
tions to compete for students transforms an input-oriented, supply-driven
education delivery system into an output-oriented, demand-driven sys-

tem. Several ECA countries (for example, Poland and Romania) have
already adopted per student financing as a core element of managing
public tertiary education institutions, shifting their focus away from
inputs and toward results.
Although most countries start the move towards per student financing
by linking financing to students enrolled, conceptually, this is not ideal.
Policy makers should encourage high student graduation rates, as well as
high rates of subsequent employment—not simply enrollment. Systems
in Denmark and the Netherlands provide an excellent model, one in
which the funding formula is based on the number of graduates pro-
duced. Other output-based funding measures are also possible, for
example, linking funds to survey results showing alumni success in the
labor market, giving institutions clear incentives to ensure that programs
are relevant to labor market needs, and aiding students in their transition
to the labor market. These steps also motivate institutions to focus on the
true outcome desired by policy makers: more qualified workers. Financing
reforms also need to be accompanied by accountability measures in order
to prevent reduced program rigor.
Performance Contracts
The introduction of performance-based budgeting can be a powerful
instrument that ministries of education (and finance) can use to guide
the sector, as it causes governments to shift from line-item to out-
come-based funding (OECD 2007c). At the same time, universities
are given greater autonomy in how they spend their budgets and are
held accountable for delivering results. Three models of performance-
based funding are currently being used in higher education (Ziegele
2009): (1) formula funding, (2) target agreements, and (3) competi-
tive funding (see table 5.2).
A recent report (OECD 2007c) reviewed the experiences of OECD
countries with performance-based budgeting and highlighted five poten-
tial benefits. First, this type of budgeting generates a sharper focus on
results within the public sector. Second, it provides more and better infor-
mation on government goals and priorities, as well as how different pro-
grams contribute to these goals. Third, it encourages greater emphasis on
planning and signals what is working and what is not. Fourth, it improves
transparency by providing greater and more useful information to the
central government and the public. Finally, it has the potential to improve
both the efficiency and the management of individual programs.
Table 5.2 Three Models of Performance-Based Funding

Formula funding Target agreements Competitive funding
Quantitative indicators Negotiation on objectives Central funds with defined
representing goals Written and signed purposes (e.g., research
Technical issues: weights, performance contract excellence, national
cost differences, formula Performance criteria priorities)
construction tailor-made to different Application procedure, often
Usual indicators: students, organizations including peer review
graduates, Ph.D.s, external “Contract” definition of Institutional and individual
income, international rewards and sanctions funding, project funding
students, etc. Ex ante definition and ex Ex ante discussion on
Ex post measurement of post measurement of performance
performance performance
Source: Ziegele 2009.
Encourage more private financial resources in tertiary education.

Mobilizing private resources can also help strengthen results-oriented
outcomes in tertiary education in the ECA region in two fundamental
ways: (1) by increasing competition, as more nimble and innovative pri-
vate providers help bring innovations to public providers; and (2) by
increasing students’ connection to the education process. When students
and parents pay tuition fees—whether to a public or a private provider—
they are generally more demanding about the quality and relevance of
the education provided.
As noted earlier in this chapter, one of the most prominent features
of the tertiary sector in many ECA countries is the presence of a vibrant
network of private educational institutions. It is thus necessary to look
at structures and levels of fees in both sectors to get a full picture of
private financing of higher education.11 Private higher education insti-
tutes in the region for the most part do not receive government funding
for core operating purposes (Turkey is the prime exception here). In
many countries, they can apply for and receive funding for scientific
research, but because their research facilities tend to be poorer than
those in public institutions, their success rate is often quite low. As a
result, private institutions are for the most part entirely self-funded via
tuition fees.
Public institutions in many countries have also been given considerable
latitude to raise funds through tuition fees. The introduction of fees has
occurred despite deep political resistance (and even, on occasion, consti-
tutional prohibitions on the practice), which is a legacy of the region’s
communist past. Generally speaking, the introduction of tuition fees has
come via a “dual-track” method in which a certain portion of students—

usually those deemed especially meritorious—are not required to pay. On
top of these students, institutions are permitted to enroll a number of
fee-paying students, although the institutional freedom to determine
both fees and the number of additional students varies widely across
the region. Thus, higher education in ECA countries has managed a
delicate political balance by both introducing the principle of fees
while retaining the principle of free tuition.
Given the significant private returns to people who graduate from
tertiary schools, however, there is an argument that the main beneficiary
of this education—the graduate—should contribute a larger share of the
costs relative to his or her peers who chose not to attend, or are not able
to attend, university. There is no single ideal level of funding for higher
education and no single ideal mix of public and private funding sources.
Different countries will make different kinds of trade-offs. Yet if ECA
countries want to simultaneously raise participation rates and increase the
quality of tertiary education, more private resources will be needed—the
cost pressures facing higher education are simply too great.
Governments could, for example, permit public institutions to collect
more revenue in return for commitments to enroll more students, or they
could facilitate the development of more private tertiary education insti-
tutions. In either case, it is critical that all institutions be given the right
to fully use their fees. It is also important that the expansion of private
institutions take place within an overall system of quality assurance, and
that appropriate financial assistance measures be put in place to assist
students in need.
Summary
In terms of the relationship between the central government and tertiary
institutions, ECA countries today can, broadly speaking, be divided into
two groups. The first group consists of most new EU member countries,
which have already expanded the autonomy of their tertiary institutions.
Their main challenge today is to strengthen a still weak accountability
structure, evident in persistent allegations of corruption in the sector and
doubts as to whether existing accountability mechanisms (such as sanc-
tions, accreditation agencies, and so forth) are sufficiently strong to induce
providers to focus on improving learning and employment outcomes
(as opposed to other goals of tertiary institutions). The second group of
countries—mostly low-income CIS countries—have maintained tight
centralized control over tertiary institutions. In order to create a more

flexible and responsive tertiary sector, these countries need to impart
more autonomy to universities, make financing a more active policy
instrument (i.e., introduce performance-based financing), and, as with the
first group of countries, introduce a range of mechanisms to strengthen
the accountability of tertiary providers.
Because strengthening accountability for educational quality is an
especially long and difficult process, ECA policy makers will need to
employ many different, complementary tools to strengthen education
providers’ accountability for results. In addition to quality assurance agen-
cies, introducing rankings, tracer studies, and standardized tests to mea-
sure student learning outcomes, together with more effective private
sector participation, will enable countries in the region to launch both
immediate and long-term quality assurance mechanisms. Counting on a
quality assurance agency to quickly establish the capacity needed to
externally monitor a rapidly growing sector is risky—this kind of agency
needs to be part of the solution, not the only response.
An essential part of strengthening accountability will be to collect
more student outcome data at the tertiary level. Without standardized
tests to measure what competencies graduates acquire and tracer studies
to establish the kinds of jobs they find on the labor market, institutions
of higher education are severely hampered in their ability to measure
their own performance or respond to the changing demand for skills.
Finally, in addition to introducing or improving performance-based
financing in the sector, countries in the region need to find ways to encour-
age more private financial resources in tertiary education. Not only will
fee-based mechanisms enable the sector to meet rising student demand
and introduce needed competition among providers, they offer an addi-
tional way to strengthen these providers’ accountability for results.
Notes
1. Given the complexity of the tertiary sector, some clarifications are needed.
First, the “providers” discussed in this chapter, both public and private, are all
institutions that offer tertiary education programs. Second, it is readily
acknowledged that tertiary education institutions serve important societal
goals that stretch far beyond that of providing graduates with employable
competencies. For instance, through their research, these institutions help
expand a society’s collective knowledge and can contribute to innovations in
the economy, giving students employable competencies in the process. And
through their political independence, universities help society hold public
officials accountable. Yet, the focus of this book—the need to create educa-
tion systems that are more responsive to labor market needs—implies that
these broader goals of tertiary institutions (and the policies related to enhanc-
ing their performance in those dimensions) are beyond its scope.
2. See a recent report on university autonomy by the European University
Association (Estermann and Nokkala 2009). The EUA report does a care-
ful job of discussing the complexities and all dimensions of university
autonomy.
3. The Bologna Process was kicked off by the signing of the Bologna Declaration
in 1999.
4. See Bologna Working Group 2005 for further details.
5. For additional information on the Tuning Project, or Tuning Educational
Structures in Europe, visit its official website at http://tuning.unideusto.org/
tuningeu/ (accessed September 2010).
6. See the online guide at www.guardian.co.uk/education/universityguide/
(accessed January 2011).
7. For a discussion of rankings, Kehm and Stensaker 2009.
8. The scorecard is the result of a regular stock-taking exercise undergone by
countries that participate in the Bologna Process. The last such stock taking
was carried out in 2009.
9. The “European Standards and Guidelines for Quality Assurance in the
European Higher Education Area” is available on the website of the European
Quality Assurance Register for Higher Education at http://www.eqar.eu
(accessed January 2011).
10. In addition to scoring countries on three indicators related to quality assur-
ance (“external quality assurance,” “student participation in quality assurance,”
and “international participation in quality assurance,” the 2009 stock-taking
exercise included a “qualitative analysis” of various aspects of internal quality
assurance.
11. Private expenditures on higher education are somewhat difficult to capture.
Data from the UNESCO Institute for Statistics on this subject is quite
inconsistent, in part because it includes income from many different sources
(see the UIS website, http://www.uis.unesco.org/; accessed September 2010).
However, it is possible to look at fee income in a relatively straightforward
way, based on accessible published sources. In most of the OECD region
(with the exception of Japan, the United States, and Mexico), for example,
fee income accrues almost entirely to public educational institutions.
CHAPTER 6
Advancing Adult Learning in ECA
Lifelong learning and adult education and training are fast becoming an
important element of education systems across the ECA region. Although
demand for labor in the ECA countries has already shifted toward
higher-order, multifaceted competencies, it has also become more unpre-
dictable, indicating the need for a market-responsive sector that can eas-
ily facilitate the reskilling of the adult labor force. The reskilling of the
work force is needed not only to compete in the global economy, but also
to address the lag effects of the transition to market economies.
Significant segments of the labor force—often comprised of less-skilled
and more disadvantaged workers—remain inactive in many areas. Given
the rapid demographic decline expected in many ECA countries, main-
taining high rates of growth will require an increase in labor force par-
ticipation and productivity.
High-quality, relevant adult learning relies on (1) sound regulation and
quality standards that promote autonomy and accountability (largely miss-
ing in all but a few ECA countries), including recognized qualifications
frameworks and certificate programs, and (2) financing incentives that
promote a competitive market and overcome the obstacles and market
failures associated with the provision of adult learning. Overcoming infor-
mation barriers for individuals and firms is central to both policy levers.
165
Unlike pre-university and university education, adult education sys-

tems are highly underdeveloped in most ECA countries. Building such
systems will require a shift away from government-defined programs
towards a well-regulated market of private and public providers that
deliver training services to both working and unemployed adults. The
preconditions for successful adult education systems are a high degree of
coordination and partnership between government agencies and the
private sector, as well as a strong voice for businesses and individuals—
that is, the demand side of training—in determining training policy and
content. Governments in the region will continue to have a role in
education and training for the unemployed, but once a solid adult
learning sector has been established, governments can then “steer” it
according to the policy framework outlined previously: by monitoring
data, promoting autonomy, and improving efficiency.
Although adult learning is of growing significance for all ECA coun-
tries, its relative importance depends on a country’s level of economic
development and demographic outlook. For some countries, such as
the EU10 and Croatia, moving quickly to build or improve this sector
will be essential; for others, such as low-income CIS countries, prob-
lems in other parts of the education system will be a greater priority in
the near term.
Building the Foundations for Adult Learning Systems

Adult education and training is a relatively new concept in the ECA
region. The experience of OECD countries suggests that developing this
part of an education system calls for a common understanding and
approach, shared strategic objectives, and functioning communication
channels among a multitude of heterogeneous players. On the part of
governments, it will require increased coordination between the minis-
tries of education, labor, and finance, other relevant government agencies
(e.g., employment services), and regional authorities. Functioning policy
coordination mechanisms are needed to give the private sector a genuine
voice in adult education, thus making the sector responsive to economic
and labor requirements. Creating such coordination mechanisms, how-
ever, takes time.
Step One: A Coherent Strategy

A typical first step in policy coordination and partnership is the elabora-
tion of a coherent strategy for adult education and training. This strategy
Advancing Adult Learning in ECA 167
creates the basis for a practical legislative and regulatory framework. The
EU, for example, has advised its member states to develop and implement
strategies on lifelong learning to advance the education and training
dimension of the Lisbon Jobs and Growth Strategy and its successor, the
Europe 2020 Strategy. The EU10 have been adopting such strategies
since 2005 (see box 6.1 for a case study).
Box 6.1
Lifelong Learning Strategy of the Czech Republic

After consultation with a wide range of stakeholders, the Czech Republic adopted
a Lifelong Learning Strategy in 2007 that includes a prominent section on
further—or adult—education. Building on an analysis of strengths and weak-
nesses, it identifies seven strategic directions.
Recognition of prior learning through the Act on the Recognition of the Results
of Further Education, with the elaboration of a National Qualification Framework,
qualifications standards, evaluation standards, and verification of professional
qualifications.
Promotion of equal opportunities in further education, particularly for disadvan-
taged groups, through financial and nonfinancial instruments, including informa-
tion and counseling services.
Development of functional literacy, including the ability to use information
technology, language skills, and other general behavioral skills.
Introduction of a system of labor market monitoring, evaluation, and forecasting
to harmonize educational opportunities with socioeconomic and labor market
developments.
Stimulation of demand for lifelong learning through elimination of financial and
nonfinancial barriers among individuals and employers, for example, through
counseling support, assistance services, and the promotion of systems of human
resource development in small and medium-size enterprises.
Measures to enhance the quality of educational opportunities, including a system
of external and internal evaluations, certification of adult education teachers, and
accreditation of educational programs.
Development of information counseling services.
The strategy is used as the guiding document for further reforms in adult
education and training in the country, as well as for program financing by the
European Social Fund.
Source: Czech Republic 2007.
The experience of EU and OECD countries suggests that ECA coun-

tries could establish multistakeholder commissions on adult learning and
task them with drafting a plan for developing the sector. Such commis-
sions typically convene national and regional stakeholders, including
representatives of key government agencies, private sector employers, and
trade unions. In drawing up a strategy, the tasks of such a commission
typically involve taking stock of existing structures related to adult learn-
ing and determining the best way to make them more relevant to the
current needs of labor markets and individual learners. In many cases, an
entirely new system may not be necessary.
Step Two: Coordination Mechanisms

Once a coherent strategy has been elaborated, coordination between
government and private sector actors is important to guide implementa-
tion. Coordination mechanisms can differ, depending on the context and
type of training. In the Republic of Korea and the United States, for
example, large multinational enterprises, together with relevant govern-
ment institutions, industry associations, small enterprises, and education
providers, have collaborated in creating “training consortia” and joint
training centers that address the strategic training needs of workers. In
the Slovak Republic and the Czech Republic, there have been similar
efforts at collaboration between large multinational companies and gov-
ernment agencies to promote adult learning in the automotive sector.
In an effort to facilitate coordination, a number of countries worldwide
have set up national adult learning bodies with broad responsibilities. The
United Kingdom, for example, has established a Learning and Skills
Council (LSC), which is responsible for overseeing nontertiary learning
targeted at those over the age of 16 years. The LSC works at both the
regional and national levels (see box 6.2). In Mexico, the National
Council of Education for Life and Work (CONVEyT) is responsible for
building a national system of education to serve adults over the course of
their lives, including clarifying the role of different providers and con-
ducting research and evaluations. Bulgaria has established a National
Agency for Vocational Education and Training (NAVET) that convenes
government, education providers, and social partners (e.g., parent associa-
tions, NGOs, chambers of commerce, etc.) to define vocational education
and training policies.
Such organizations are, however, unlikely to be successful if the rele-
vant governmental ministries are not prepared to fully cooperate with
them and provide them necessary support (OECD 2005c). Creating new
Box 6.2
Policy Tools for Advancing Adult Learning

in the United Kingdom
The Skills Alliance, a ministerial-led group of adult learning stakeholders, includ-
ing employers, provides oversight of the Skills Strategy in England. Employers
offer input to the Alliance through their representatives, together with the Skills
for Business Network and the National Employment Panel (NEP). The provision of
training is organized and funded by Learning Skills Councils (LSCs) at national,
regional, and local levels in partnership with Regional Skills Partnerships, col-
leges, providers, and other key stakeholders. Providers include publicly funded
bodies, nonprofit agencies, and private education providers.
The LSCs allocate funds based on “expected demand.” The planning process
involves taking stock of the existing learning environment, drawing on different
sources of labor market intelligence, and dialoguing with key partners, including
employers and providers. Implementation involves the securing of training
through open bids and closed negotiations managed by LSCs. In an effort to
make the system less supply driven, an independent review of the U.K. adult
learning system, the 2006 “Leitch Review,” laid out the following principles for
adult education and training:
(1) demand-led funding, that is, routing public funds for adult vocational skills
training through demand-led channels, thus ending supply-side planning of
skills provision;
(2) strengthening the voice of employers by rationalizing the number of bodies
that articulate the views of employers and creating a single Commission for
Employment and Skills; and
(3) providing economically valuable skills by relying on reformed and relicensed
Sector Skills Councils to ensure that vocational qualifications reflect the skills
valued by employers (Leitch Review of Skills 2006).
Similar to other EU countries that have introduced National Qualification Sys-

tems, the United Kingdom established the National Vocational Qualification
(NVQ) system to recognize a broad spectrum of vocational skills that can be
acquired both on and off the job. Most critically, the NVQ system recognizes prior
learning through an assessment process. The modular system allows participants
to learn at a pace convenient to them. During the period 1988–2003, approxi-
mately 4 million NVQ qualifications were awarded, with the majority at Level 2 of
Box 6.2 (continued)
the system (Dearden, McGranahan, and Sianesi 2004). Yet despite the growth in
these qualifications, wage returns to NVQ Level 2 qualifications have to date
largely been low or nonexistent, with a few exceptions (Dearden, McGranahan,
and Sianesi 2004).
In addition, various projects provide information to workers to keep them bet-
ter informed about market demands and the returns to training. A number of pro-
viders also offer advice to workers, particularly those who are unemployed or
qualify as low skilled. These efforts include Jobcentre Plus work to identify the skill
needs of Jobseeker Allowance recipients who have been unemployed and inac-
tive for six months; Learndirect’s personalized advice, information, and guidance
tailored to the needs of adult clients at the skill level NVQ 2 or below; and the com-
munity outreach of Learning Ambassadors, local volunteers who have themselves
undergone adult training, to identify and work with people in similar situations.
Individual Learning Accounts (ILA) in the United Kingdom provide funding to
individuals with the lowest skill levels, making training accessible to a group that
otherwise would not participate. ILAs make a maximum sum of between £100
and £200 available to low-skilled individuals, who may spend it on a qualification
and provider of their choice. Individuals and their employers are also encouraged
to contribute to ILAs. Opening an ILA comes with a guidance package that targets
individuals who are least able to borrow from private financial institutions. In
2000, England implemented one of its largest ILA projects, with 1.4 million par-
ticipants, but it was later closed due to irregularities. The project was found to
suffer from significant deadweight loss, with more than half of participating indi-
viduals stating that they would have sought training without their ILAs (OECD
2005c). Subsequently, however, both Scotland and Wales implemented compa-
rable schemes and since 2007, England again began to pilot a similar project.
Lastly, the government’s Train to Gain Program helps employers identify train-
ing needs and then tailor corresponding training packages which are jointly or
fully financed by the government. The program offers financial incentives to
workers with paid time off and subsidies to employers to compensate for trainee
wages. This intervention targets all market failures, from financing training to over-
coming the time preferences of employees, who worry about the opportunity
cost of training. The Train to Gain initiative was first piloted in 2002–2006 and
covered 30,000 employers and 250,000 employees. It has now been mainstreamed
across the country. Initial evaluations indicate a small positive impact on the take-
up of training for both employees and employers (Abromovsky et al. 2005).
Sources: Abromovsky et al. 2005; Dearden, McGranahan, and Sianesi 2004; OECD 2005c; Leitch Review
of Skills 2006.
bodies carries the risk of adding additional layers of bureaucracy; and

some countries, such as the Czech Republic, have managed to enhance
cooperation between ministries without forming new agencies.
Given the varying needs across an economy, sector- or industry-level
councils can also be key components of successful adult learning systems,
especially for defining the standards of vocational training and developing
national qualifications frameworks (see below). Permanent regulatory
bodies, on the other hand, help ensure that qualifications remain relevant
to labor market requirements over time. The United Kingdom, for exam-
ple, currently relies on a network of employer-led Sector Skills Councils
(SSCs) to coordinate different market stakeholders and develop quality
standards. The SSCs are monitored by the Sector Skills Development
Agency, which also ensures that cross-sectoral skills are adequately cov-
ered by U.K. licensing systems (United Kingdom 2003).
Building on the United Kingdom’s example, the Czech Republic has
formed Sector Skills Councils with the involvement of social partners.
The government convened the human resource experts from leading
companies to help prepare a national occupations framework, which,
in turn, laid the foundation for a national qualifications framework.
In Croatia, sector councils—reporting to the Agency for Vocational
Education—have recently been established to assess skills needs in spe-
cific vocational sectors and contribute to setting occupational standards
(World Bank 2009a).
Building Adult Learning Systems in the ECA Region: Uneven Progress

Although many ECA countries, particularly the EU10, have made
progress in developing the kinds of coordination systems described above,
several challenges remain. First, the voice of labor market stakeholders in
decisions on adult learning continues to be weak relative to that of the
government. For example, in Hungary, education and enterprise repre-
sentatives were involved in developing vocational qualifications and cur-
ricula, but employers viewed the qualifications as closely modeled on
those of school education, too supply-driven, and insufficiently forward
looking (Gunny and Viertel 2007). A large sample of top managers
judged that the allocation of Hungary’s employer-based training funds
was influenced most by government institutions, followed by training
institutions, and finally, by employers’ associations. Not surprisingly, they
felt that this order should be reversed (Godfrey 2003).
Second, coordination challenges, particularly across ministries, persist.
In Hungary, for example, coordination between the ministries of labor
and education is limited, evidenced by the fact that they operate separate
adult education programs.1 In Poland, the ministries of labor and educa-

tion are statutorily required to cooperate on adult learning initiatives,which
must be approved by the Council of Ministers. In practice, however, there
has been little coordination (OECD 2005d).
The evidence for adult learning interventions to date is either lim-
ited or mixed; the success of their effectiveness depends on context,
specific design features, and to whom they are targeted. Building adult
learning systems thus requires carefully examining the successes and
failures of existing international policies and programs. Moreover, gov-
ernments in the region will benefit from rigorous monitoring and
evaluation, including impact evaluations, of such programs to ensure
that they meet desired objectives.
Promote Autonomy and Accountability

of both Public and Private Providers
Autonomy
Adult education is perhaps most effective when it functions as a market
for training. If informed well, consumers will demand programs of high
quality that are relevant to their employment aspirations. If autonomous
and agile, providers will respond to this demand and offer quality pro-
grams in a competitive environment. If providers are held accountable for
the quality of the training provided, programs will meet the highest and
latest standards. If, for example, a computer training provider delivers
programs of limited quality and relevance, consumers will simply shift to
another provider.
Arguably, private providers are already autonomous in their deci-
sions regarding market entry and exit, as well as regarding investments
and products. The same is not necessarily true for public providers if
they receive institutional funding from the government and are bound
by government rules. In order to truly compete in the adult education
and training market and deliver useful, state-of-the-art programs, pub-
lic providers also require autonomy to respond to changes in demand
and other competitive pressures. For instance, to compete successfully,
vocational training providers and tertiary education institutions will,
similar to private providers, need to be autonomous in their decisions
regarding market entry and exit, as well as regarding investments and
programs. As an example, the Czech Republic is encouraging public
vocational education and training schools to compete in the adult
training market.
Accountability
Information is the key to accountability in adult education and training.
Functioning accountability systems require government regulatory
intervention to ensure that (1) adequate information is available for
consumers of such education (so that they may make informed deci-
sions), and (2) the suppliers of adult education and training programs
maintain the quality and relevance of their program offerings. Thus,
policies that set standards for and ensure the quality of adult learning
opportunities both foster supply and stimulate demand. For example,
national qualification frameworks can help to stimulate demand by
allowing workers to obtain certified qualifications that prove their com-
petencies to employers—and address a key information constraint. They
can also make continuing vocational education and training (CVET) and
retraining of unemployed workers more strategic and relevant to eco-
nomic needs. Clearly, the regulation and certification of education pro-
viders—both public and private—and their programs is a key element of
developing an adult learning system.
National qualification frameworks (NQFs) are being developed in
many ECA countries to establish coherent quality standards for educa-
tion and training systems and recognize prior learning, including that
acquired through formal, nonformal, and informal learning methods. For
many countries in the ECA region, NQFs—at least for the tertiary level—
are a mandatory component of the Bologna process (see chapter 5). The
development of such frameworks is, however, complex. NQFs require a
strong legislative basis, a high degree of coordination and administrative
capacity, and a great deal of time—countries such as Ireland and the
United Kingdom have spent years developing and refining them (Gunny
and Viertel 2007; Ireland 2003; United Kingdom 2003). Moreover, they
are no panacea, and it is important to set realistic expectations as to what
they can achieve in conjunction with other policy levers in adult educa-
tion and training. A recent assessment of the experience with NQFs in
16 countries worldwide showed that they may not always result in the
better matching of demand for and supply of adult education and train-
ing, nor in better training outcomes (Allais 2010). The success and failure
of NQFs appear to depend, at least in part, on the extent of the construc-
tive involvement of key stakeholders—and their willingness to eventually
make use of frameworks devised.
As noted in the previous section, employer-led, sector-level councils
often play a critical role in developing standards. For example, in the
Czech Republic, sector skills councils have started to develop national
occupational frameworks for each occupation, together with their

required key competences and qualifications. National qualifications
standards are also being developed based on this information. By early
2009 more than 160 qualifications standards were in place, with an addi-
tional 250 yet to be defined. According to the Ministry of Education,
Youth, and Sport, the eventual number of qualifications could reach
between 1,000 and 1,200. Given the magnitude of NQFs, it is practical
for countries to begin defining standards for a few important occupa-
tional sectors before working towards a broader qualifications framework.
The Kyrgyz Republic, for example, has begun piloting a qualifications
system for the tourism sector as part of developing an NQF (ETF
2007).
In addition to setting appropriate standards and coordinating relevant
stakeholders, some countries also face social barriers when implementing
qualifications systems. For example, members of the formal education
system sometimes believe that they are the sole legitimate providers of
education and training, making them averse to systems that accredit for-
mal and informal learning results on an equal basis (Hungary 2006).
Similarly, a shift in thinking is also required on the part of employers,
who, when hiring workers, will need to be open to considering not only
recognized education degrees, but also partial certified qualifications—
that is, an individual’s job qualifications that do not necessarily stem from
formal education degrees. Again, the actual use of NQFs is likely to
depend on the extent to which key stakeholders have been an active part
of their formulation.
In addition, some adult learners are intimidated by the process of
obtaining qualifications. Although this does not appear to be a major bar-
rier to adult education and training in many ECA countries, it may hinder
less educated workers in particular. In order to attract these workers,
regulators and providers can design learning programs in small units and
ensure that testing methods are approachable. For example, computer-
based testing, multiple assessment opportunities, and small group learn-
ing have all contributed to increasing the participation of low-skilled and
immigrant workers in the United Kingdom as part of the London Open
College System (United Kingdom 2003).
Quality standards and accreditation systems are also important for
ensuring the quality of adult learning providers, especially those that
receive public funding. Several OECD countries have developed quality
assurance, inspection, and accreditation systems that can serve as models
for ECA countries. For example, Denmark, Portugal, and Spain have
developed institutes that evaluate adult learning programs with the aim
of ensuring quality. The United Kingdom has established an Adult
Learning Inspectorate that assesses hundreds of publicly funded programs
every year and publishes the results; those that underperform often lose
funding (OECD 2005c). Quality assurance systems in adult education
and training can also involve the use of self-evaluations by providers
alongside external evaluations, as in the case of Slovenia. In the Czech
Republic, the Ministry of Education, Youth, and Sport certifies retraining
programs for the unemployed that are financed by the Labor Office. The
ministry is also beginning to pilot a quality assessment system and plans
to institute a star rating scheme for quality based on a simple set of crite-
ria, rather than complicated quality standards. If successful, this could be
an interesting option for less advanced countries.
In ECA countries, programs that assess the quality of adult education
and training may play an important role in helping adults, particularly
those with little experience of these programs, make decisions. As noted
in box 6.3, cost is a major barrier to adult education and training in the
region: Polish providers have determined that learners are likely to
choose a program based only on price, irrespective of quality (OECD
2005d). For providers that do not typically receive public funding, qual-
ity seals are an attractive means of assessment. Even though such seals
rely on voluntary participation, providers have the incentive to meet
standards in order to demonstrate the quality of their product to the
market. Both Germany and Austria have developed quality seals. In
Austria, the initiative was originally supported by nonprofit providers
and in Germany by market providers, but both receive funds from the
government (OECD 2005c).
Ensure the Efficiency of Sector Financing

Financing for adult education and training programs should largely be
private, certainly for most continuing vocational education and training
programs. However, public funding will be needed to address market
failures in the sector, provide incentives to encourage the participation of
underrepresented groups (e.g., older and less skilled workers), and finance
programs for the unemployed. Governments already play a role in the
financing of adult education and training in ECA, funding retraining for
the unemployed through public employment services and providing
incentives for private investments. The existence of market failures in the
sector presents a rationale for continued and, occasionally, expanded
Box 6.3
Market Failures that Impede Adult Learning

Private investments lie at the heart of adult education and training in most
countries, but governments can create an enabling environment for private
investment through careful regulation that facilitates the flow of information
and ensures educational quality. Firms and workers—not governments—
should drive adult training decisions, as their main motivation for investing in
training is the expectation of greater profits and higher wages, respectively.
Governments, conversely, recognize both the social returns and overall eco-
nomic benefits of increased human capital. Accordingly, it is in their interest to
provide a certain amount of financing to overcome market failures associated
with the provision of adult learning, including externalities, credit market fail-
ures, and information failures.
The economic literature identifies a range of market failures in adult educa-
tion and training, suggesting that the participation of individual workers in such
training may be below the socially optimal level, despite its considerable returns.
In other words, individual workers and company managers are making rational
and optimal decisions about training given the constraints that they face, but
the sum of these decisions represents a less than optimal outcome for individual
countries. First, externalities may hold back training. Economic theory suggests
that because firms do not enjoy the full benefits of training—some of it will
accrue to future employers—they may invest less in it. For example, the higher
the risk that a worker will leave a company soon after training (i.e., the “poaching”
of trained workers), the more firms will be cautious in investing in this area. While
no evidence from firm surveys in the ECA region documents an externality bar-
rier, anecdotal evidence suggests that it is an important concern for employers
in the region, particularly in those countries that, until recently, faced a very tight
labor market.
Second, individuals and firms may also be unable to generate the necessary
funds to finance education and training (credit market failure). For example, firms
may be unable to raise financing in the capital markets themselves, especially
small firms with limited physical collateral. Survey data from the EU10 suggests
that the cost of training is presently an important barrier to more training (see
figure B6.3). Third, workers and firms may opt out of training because they lack
information, for example, about the availability and quality of training providers.
Lack of recognition or certification of learning will also reduce workers’ interest in
Box 6.3 (continued)
Figure B6.3 Barriers to the Expansion of Adult Education and Training in

the ECA Region
70 a. Training was too expensive or respondent could not afford it

% of respondents answering
60
that this was a constraint
50
40
30
20
10
0
A
N
T
P
U
P
L
R
R
T
C
U
E
R
ES
PO
ES
AU
CY
SW
SV
IT
UK
BG
NO
FI
GR
LV
LT
DE
HU
other EU EU 10
b. Access to information on learning possibilities, all types of education

80
% of respondents participating in
informal or nonformal education
SWE
70
60
FIN
50 UKR
DEU SVK
40 CYP AUT
BGR
LTU FRA
LVA
ESP
30
20 ITA POL
GRC
10 HUN
0
0 10 20 30 40 50 60
% of respondents reporting that information regarding
training opportunities is available
Sources: Authors’ calculations based on AES data (2007).

Note: See “Abbreviations” for country abbreviations and groupings.
training because they cannot be sure that they can demonstrate their raised
productivity to employers. Conversely, employers and workers in developing
economies may not be fully aware of the productivity returns from training due
to lack of information about the type and area of training, as well as a lack of
managerial skills in small and medium-size enterprises. Survey data from ECA
confirm a positive correlation between the incidence of training and the avail-
ability of information (see figure B6.3, panel b), with the information barrier par-
ticularly acute for less educated workers.
Source: Authors.
government investment in adult learning in the region (see box 6.3).

Government financing is unlikely to make a difference, however, unless it
is used strategically to support better partnerships and overcome the
disconnect between different market players.
Public funding will be most effective if it provides incentives that
promote quality, empowers the demand side, and follows competitive
principles. To make government financing effective, ECA countries will
need to focus on outcomes, rather than inputs, thus shaping a competi-
tive market for adult education and training. For example, government
funding for training programs for the unemployed can be made com-
petitive and outcome-based. State funding can also address market fail-
ures that result in lower average participation rates, or lower participation
by certain groups of workers, by offering the private sector more incen-
tives to invest in adult education and training, such as voucher programs
and tax breaks (see the following section). The scope of government
financing for adult education and training will vary by the type of training
and target group. Because retraining programs for the unemployed,
including the underdeveloped field of “second-chance” education, gener-
ally have social returns, they are generally financed by state resources.
Conversely, CVET and education for the employed generate substantial
private returns to firms and individuals, and should therefore be largely
financed by these groups.
Public financial incentives for CVET have been rare in the ECA
region, but are likely to become more prominent as economies become
more technologically advanced and the need for “upskilling” becomes
more prominent. Again, it is important that countries in the region create
the right incentives and focus on strengthening demand. As described
earlier in this chapter, governments in many OECD countries use public
financing both to promote more private investment in training and to
overcome market failures. Two incentives for firms found in many OECD
countries are tax deductions for training costs and payroll training levies.
Some ECA countries have also begun experimenting with these tools. In
addition, a number of EU10 countries have begun utilizing financing
from the European Social Fund (ESF) to create grant programs for firms,
while others have used demand-side financing, such as vouchers, to target
individual workers to participate in training.
Offer Tax Deductions to Firms that Offer CVET

Allowing firms to deduct the costs of training from their taxes is one
of the most common forms of training incentives for firms. Such tax
deduction schemes have been implemented in some countries in the ECA

region, particularly in the EU10 and the Russian Federation, and have
been a useful first step for stimulating demand for training. Tax deduc-
tions for training costs are relatively simple for governments to administer
and for firms to use, as they rely on existing tax systems. Although tax
deductions generally benefit only firms that are profitable, this bias can be
managed by allowing deductions to be spread over a period of years.
However, tax deduction systems can also result in deadweight loss and
higher administrative costs. Deadweight loss occurs when individuals and
firms take advantage of tax deductions even though they would have
offered training without them. For example, large firms are generally
most likely to take advantage of deductions and are more prone to use
funds to train highly skilled workers (Cedefop 2009). This reality is why
several OECD countries employ more complex, multilevel tax deduction
systems that target financial incentives to those firms and individuals who
face barriers to training. While such targeting is attractive in theory, it also
increases administrative costs, invites manipulation by firms and makes
the system more difficult to use. In an effort to reduce deadweight loss,
for example, the Netherlands experimented with targeted tax deductions
but decided later to abolish them because they were ineffective and too
complex (OECD 2005c). However, despite these challenges, tax deduc-
tion schemes can be a valuable tool for giving an initial boost to adult
education and training, particularly in environments where few firms and
individuals currently provide training.
Consider Payroll Levies and Other Grant Mechanisms

Grant systems based on payroll levies can also create an effective funding
source for training and allow for greater targeting of training funds.
However, such systems can be challenging to implement. This type of
grant system levies a payroll tax (typically of 0.5–2.0 percent) and dis-
burses the funds collected to firms for training purposes, either in the form
of reimbursements or grants (Dar, Canagarajah, and Murphy 2003; OECD
2005c). There is promising evidence from OECD countries that well-
designed training levies can increase training rates, particularly for
medium-sized firms (OECD 2005c; Tan 2001). By allowing grants to be
targeted to firms and individuals who are less likely to seek training in the
absence of a financial incentive, levy grant systems can in theory also
be designed to limit deadweight loss. However, to implement an effective
levy grant system, countries need to consider the preparedness of firms to
participate in such a system, management of the funds, and ways to ensure
the actual participation of small and medium enterprises (SMEs) in train-

ing programs. Careful communication and coordination with the enter-
prise sector is crucial, as such systems can be problematic in situations
where firms complain of high payroll taxation and believe that they will
have a low chance of successfully accessing and utilizing training funds.
Even in countries where payroll levy grant systems have had a gener-
ally positive impact on raising training participation, noncompliance rates
are often a concern, particularly in less advanced countries and among
small firms (Dar, Canagarajah, and Murphy 2003). Most existing levy
grant systems rely on a central agency that manages the funds and awards
them to firms—a role that can be filled by a government or social part-
ners (Ok and Tergeist 2003). To be effective, levy grant agencies require
strong capacity in administration, quality control, and training policy.
Levy grant systems can also be managed by a firm itself; Poland, for
example, has set up a combined training fund and subsidy based in indi-
vidual firms, thus avoiding a central management structure. If a firm com-
mits at least 0.25 percent of its total wage bill to a training plan designed
in cooperation with employee representatives, the government will sub-
sidize the training (OECD 2005e).
Despite the training benefits for medium-sized firms, one of the draw-
backs of levy grant systems is that they tend not to benefit either SMEs or
very small firms. The most effective way to implement a levy grant pro-
gram may thus require additional targeting of SMEs and exempting the
smallest firms (e.g., those with fewer than 50 employees), which are pro-
vided a different type of assistance. (For specific barriers to SMEs that are
not alleviated by simple financial incentives, see box 6.4.) The Human
Resource Development Fund (HRDF) in Malaysia is an example of a suc-
cessful payroll levy program that alleviates the additional constraints faced
by SMEs and reduces the administrative burden. For example, rather than
requiring SMEs to develop their own training plans, established training
providers are used to offer time-tested and preapproved training programs.
This feature also helps stimulate supply by giving training providers the
opportunity to market well-defined programs (Tan 2001).2
Grant schemes play an important role in many ECA countries, par-
ticularly in the EU member states, which have access to grants from the
ESF. They use these grants to support lifelong learning programs in
line with the EU Lisbon Jobs and Growth Strategy and its successor, the
Europe 2020 Strategy. These funds, however, sometimes bolstered the
supply, not the demand, side of adult education and training. Instead of
providing workers with vouchers, they have often been channeled to
training institutions. Some countries have utilized general budget funds
Box 6.4
Supporting Small and Medium Enterprises

to Participate in Training
Small and medium enterprises often face a variety of constraints that make it dif-
ficult to provide and take advantage of training, such as obsolete technology, lack
of access to finance, limited administrative capacity, poor management skills, and
an inability to use training effectively. Programs that alleviate credit and financial
constraints, however, may not be sufficient to stimulate more worker training—such
programs must be designed to address additional constraints as well. The Integral
Quality and Modernization (Spanish acronym CIMO) Program in Mexico was devel-
oped when it became clear that training subsidies alone were not effective in in-
creasing worker training in many SMEs. Firms who participate in the program un-
dergo a diagnostic process that identifies their production methods and assesses
workers’ skills needs and other firm constraints. CIMO then organizes clusters of
enterprises (where possible), links firms to outside providers, and subsidizes up to
half of the cost of training per firm—a set of policies that is likely to be more effi-
cient than having the government provide the training directly.
An evaluation of the CIMO program suggested positive impacts on training
participation rates and enterprise performance. In sectors with many SMEs, par-
ticularly where they partner or act as suppliers to larger firms, programs to pool
resources are promising. For example, the Korean government has supported the
development of large enterprise-led training consortia in response to the low
take-up of payroll tax-funded training grants by SMEs. Large enterprises have
developed training consortia in response to skilled labor shortages and low-quality
partner organizations and suppliers. The consortia—which have had very high
participation rates—can capitalize on training resources that already exist in large
enterprises, benefitting from higher-quality inputs and more efficient relationships
with partners. They receive some direct subsidies and also benefit from a higher
uptake of existing training grants. In the ECA region, the Czech Republic is launch-
ing a similar SME retraining voucher program funded by the European Social Fund.
The program will promote professional retraining, pooling resources from enter-
prises, municipalities, and the ESF.
Source: Tan and Lopez Acevedo 2005; OECD 2005c; OECD 2004a.
for grant schemes, especially to fund services for the unemployed.

Bulgaria, for example, is now operating a grant cofinancing scheme to
encourage in-service training of the workforces of private companies.
Careful evaluations could help determine the effects of these efforts.
Alternative Financing Mechanisms: Provide

Support Directly to Workers
Demand-side programs that offer incentives directly to workers are an
alternative or supplement to supply-side programs that finance adult
education and training providers. Since financial and time constraints are
often cited by workers as hindrances to training (see box 6.3), programs
that directly address these problems can be beneficial. These programs
build on the individual motivation to learn, which may increase the
chances of learning success, and also have the advantage of directly help-
ing workers with the greatest skills deficits.
The simplest and most widely implemented demand-side program is a
tax incentive that allows workers to deduct the cost of self-financed train-
ing. Such programs more effectively benefit targeted individuals than do
tax incentives for firms; however, individual-based tax incentives only
reach those workers who earn wages high enough to pay income taxes.
An alternative is training leave, which provides workers reduced working
hours or breaks from full-time employment with some degree of com-
pensation, to enroll in training programs, combined with the guarantee of
a job upon completion of the training. This type of program is used in
many OECD countries and takes various forms in terms of targeting,
length, and the level and source of financial compensation (see OECD
2005c, chapter 3).
Grants and vouchers are increasingly popular programs that provide
incentives for workers to seek education and training. The advantage of
these programs is that different levels of subsidies can be offered to dif-
ferent types of workers, based on need. For example, education and train-
ing voucher programs in Austria offer more generous support to older
workers, while the Adult Learning Grant (ALG) in the United Kingdom
offers an “allowance” that varies in size according to the means of the
recipient (IFF Research 2008). Voucher and grant programs oriented
towards shorter courses have been designed to cover only education and
training costs, while programs geared towards higher qualifications must
usually offer subsidies that provide living expenses in order to be effective
(see OECD 2005c).
Like several other new EU member states, the Czech Republic is using
financing from the ESF for a demand-side voucher program called “Pivo.”
The program targets both employed and nonemployed individuals aged
25–64 years to finance education and training in languages, information
and communication technology (ICT), and entrepreneurship skills—all
skills that have been identified as underdeveloped among the Czech
workforce. However, training vouchers do have limitations. Recent evalu-

ation evidence from the United States, for example, indicates that their
effectiveness may be hampered by faulty information, as when up-to-
date data on local skill needs are not available to either training providers
or participants (particularly disadvantaged and low-skilled participants)
(Barnow 2009). In addition, the effectiveness of vouchers can be affected
by a limited supply of quality training providers, at least in more remote
geographical locations—though the existence of vouchers can help
stimulate the training market.
As noted earlier in this chapter, Individual Learning Accounts (ILAs)
are another form of support for individual workers. Individually desig-
nated bank accounts that can receive financing from a firm, an indi-
vidual, or the government, ILAs encourage investments in adult
learning but can be onerous to administer. Like other demand-side
financing schemes, ILAs are appropriate only where it is possible to
safeguard against abuse and fraud. Indeed, England abandoned a large
ILA program because of this very concern. Nevertheless, ILAs are
appealing because they often rely on matching funds, sometimes in a
(tax-sheltered) savings account. Although several OECD countries
have found them difficult to implement, some of their large employers
have independently started implementing ILAs to encourage employ-
ees to do training. Providing tax relief for firm-based accounts, where
they exist, could also be a helpful way to encourage education and
training in ECA countries.
Continued Government Role in Retraining

and Education for the Unemployed
In most ECA countries, retraining and education programs for the unem-
ployed are largely publicly funded and represent a significant share of
government spending on active labor market programs. The record on the
effectiveness of such training is mixed, but there is evidence that they can
have positive impacts on employment if well designed, especially over
the medium term (Betcherman, Olivas, and Dar 2004; Card, Kluve, and
Weber 2009). Access to training for the unemployed in the ECA region
is limited (see chapter 2), yet its importance is set to grow in light of
shrinking populations and the increasing need to expand labor force par-
ticipation. Retraining programs can make a crucial contribution to pro-
moting employment by strategically reskilling laid-off workers for new
jobs—an important agenda in the wake of the recent economic crisis.
When focused on basic skill gaps, these programs can also promote social
inclusion (see box 6.5).
Although many ECA governments have long funded industrial
training institutes that trained or retrained unemployed adult workers,
in most instances, these institutes were not always designed to meet the
needs of the modern work force. Governments today would be better
Box 6.5
Second-Chance and Remedial Education

The demographic decline being experienced in most ECA countries suggests that
every person of working age needs to be brought into the labor force. This means
that, in addition to the equity imperative, there is a growth imperative for greater
skills. Many unemployed and economically inactive people in the ECA region face
educational disadvantages as a result of leaving school early or a failure to acquire
basic skills. Second-chance and remedial education and training are promising
tools for addressing basic skills deficits that prevent adults and young adults from
entering the labor force and becoming sustainably employed. These programs
often involve basic literacy and numeracy training. Overall, second-chance edu-
cation programs remain limited in many ECA countries, even though basic skills
deficits often are a key barrier to the labor force participation of disadvantaged
workers. Some countries in the EU10, such as Bulgaria, have recognized that lit-
eracy is a concern for a small but significant portion of the disadvantaged popu-
lation, including the Roma minority, and have begun to extend their literacy
programs in response.
Opening new pathways to employment through this type of education
requires developing new outreach and delivery mechanisms. Employment ser-
vices in many ECA countries have a poor record of effectively serving the most
disadvantaged.a Basic skills gaps such as functional literacy are typically not even
considered when identifying active labor market interventions for the registered
unemployed, even if these people are highly disadvantaged. Second-chance
education and training are thus ideally part of employment activation programs
for the disadvantaged, long-term unemployed, and welfare-dependent adults.
However, these programs require employment services and social welfare offices
to integrate their efforts in partnership with NGOs and community-based organi-
zations—both in order to reach beneficiaries and to deliver tailored support to
them.
Box 6.5 (continued)
Second-chance education and training programs often involve behavioral

and social skills training—especially for young people who have failed to enter
the workforce and for other long-term unemployed who have attained some
basic level of education. Many such individuals are from disadvantaged back-
grounds and lacked the opportunity to develop the social and behavioral skills
needed to effectively hold down a job. For example, Czech employers cited
irregular job attendance and low work motivation as reasons for not hiring
(especially young) Roma (World Bank 2008a). It is thus critical that job training
for such groups take such concerns into account. The youth-oriented Jovenes
program in Latin America, which was deemed successful in short-run evalua-
tions, is an example of a program that successfully combines training in voca-
tional skills with a curriculum aimed at improving communications, personal
relations, and self-esteem.
a. For evidence from the Czech Republic, see World Bank (2008a).
off by contracting out retraining services through a competitive bid-

ding process, rather than relying on these training institutes. In this way,
they would both foster the restructuring of formerly public providers
and trigger the market entry of new, private providers. As a large pur-
chaser of training services for the unemployed, a government can
implicitly set the rules of the overall adult education and training mar-
ket and thereby foster more private investment, including in CVET, if
public financing is following competitive rules. The Czech Republic
followed this approach in shaping its training market, which resulted,
deliberately or not, in substantial market entry and a largely private-
provider market.
Retraining programs in ECA countries can become more effective if
they are designed to meet identified labor market needs; center on the
needs and capabilities of individual job seekers; and are well linked to
other active labor market programs, such as job referral services. Effective
training and requalification efforts would also benefit from regular labor
market assessments, including analysis of vacancy data, to gauge changes
in labor demand and ensure that jobseekers become retrained in voca-
tions that are actually in demand on the labor market.
As valuable as such analysis is, it can be difficult to implement in
practice. In the Czech Republic, for example, labor offices conduct
regular assessments of vacancies at the regional and municipal levels.

However, the actual provision of training does not always respond to
such assessments, as certain training programs are procured in bulk by
the labor office. In Poland, it has been argued that regional labor offices
lack the resources and training to conduct labor market assessments.
Clearly, capacity building in this area is a critical to ensuring that such
assessments take place (OECD 2005d).
When contracting out training to private providers, output-oriented
incentives such as performance-based contracts can boost the relevance
and effectiveness of the training. Turkey has had good experience with
contracts that include built-in incentives for private providers to achieve
high job placements rates. Until 2008, the Turkish Public Employment
Service (ISKUR) followed this course to ensure that there was an actual
demand for the skills that providers offer in their training classes.3 Recent
ISKUR data provided to the authors indicate significant differences in
placement rates for in-class training programs between providers that
have employment incentives (i.e., employment guarantees) and those
that do not, highlighting the importance of contract design in placement
intermediation (see figure 6.1).
Figure 6.1 Employment Placement Rates by Type of Retraining, Turkey, 2008
90
80
70
60
percent
50
40
30
20
10
0
ar m nd
ar m ing
ar m ing
gu ploy I fu
te t
te t
te t
gu ploy ain
gu ploy ain
an en
an en
an en
e)
e)
e)
U
r
m t
m t
em ing
(e lass
(e -job
(n rain
-c
e
-th
in
t
ss
o
on
- cla
in
Source: Staff calculations based on administrative data for 2008 provided by the Turkish Public Employment
Service (ISKUR).
Priorities for Adult Education and Training Systems

in ECA Countries
Given that the development of modern adult learning systems is a complex
and lengthy task, countries in the ECA region need to carefully prioritize
their efforts in this sector. In many cases, improving the quality and equity
of other levels of education may be of greater importance in the near term.
Figure 6.2 suggests priorities for countries in the region based on their GDP
per capita (as a measure of economic development) and demographic out-
look. The figure groups ECA countries into three broad groups:
(1) Advanced economies facing a demographic decline that compete in

highly competitive markets (i.e., the new EU member states, Croatia,
and Russia);4
(2) Less advanced economies facing a demographic decline (i.e., many
countries in South Eastern Europe plus several rapidly aging CIS
countries); and
Figure 6.2 Adult Education and Training Priorities in the ECA Region
group 1: less
advanced, with group 2:
20
demographic crisis: advanced, with
consider demographic
UKR BGR crisis; priority
BLR RUS
GEO
10 ROM LVA
LTU
percentage of population decline 2010–30
MDA HRV
H V
BIH POL N EST
HUN
ARM CZE SVN
SV
MKD SVK
0 SLB
MNG
ALB KAZ
–10 AZE
KGZ
–20 TUR
UZB
group 3: less
–30 advanced, with no
demographic crisis:
TJK
no priority
–40
0 5,000 10,000 15,000 20,000 25,000 30,000
GDP per capita (2007 US$)
Source: Staff calculations based on United Nations 2006.

Note: See “Abbreviations” for country abbreviations.
(3) less advanced economies facing a demographic expansion, consisting

of the majority of the Central Asian countries along with Albania.
Turkey and Kazakhstan do not fit well into any of these categories, but
it can be argued that their priorities are the same as those of group 1.
Table 6.1 provides a rough policy framework for adult education and
training in ECA countries, based on their respective economic and demo-
graphic indicators.
Advanced Economies in Demographic Decline

region that are facing a demographic decline. Their challenge is to ensure
that existing coordination mechanisms function well and that regulation
and financing are used to facilitate the emergence of a private sector-
oriented adult education and training market. As shown in chapter 2,
some EU10 countries are relatively advanced and on par with many older
EU member states in terms of training participation. The Europe 2020
Strategy places a strong emphasis on lifelong learning and has served as
an engine for policy change and the programming of EU funds. These
funds, in turn, have stimulated greater promotion of adult education and
training. The strategies in these countries are in place, supported by
multistakeholder coordination bodies, and legislation is being introduced
alongside regulatory regimes and new financing schemes.
The challenge in the EU10 and Croatia is to ensure that the demand
for education and training emanating from the labor market can be
served by flexible, high-quality programs. Within this group, certain
countries, such as the Czech Republic and the Slovak Republic, have
high training participation rates and a culture of adult education and
training has taken root. Here, the task is to make training more effec-
tive and strategic, improve quality through autonomy for providers and
accountability mechanisms, carefully evaluate outcomes, and use gov-
ernment financing in a tactical way. Across the region, given their
importance, special attention should be given to encouraging training
in small and medium enterprises. A particular challenge will be ensur-
ing that less educated, older, and disadvantaged workers can participate
more in education and training, regardless of whether they are
employed, unemployed, or outside the labor force.
Turkey and Kazakhstan might also consider following this agenda for
adult education and training. Although neither country fits well into the
country groupings used in this chapter, both countries have expanding
Table 6.1 Recommended Policy Framework for Adult Education and Training in
the ECA Region
Advanced Less advanced Less advanced
economies with economies with economies with
demographic demographic no demographic
challenge challenge challenge
ARM, BIH, BLR,
GEO, MDA, MKD, ALB, AZE, KGZ,
EU10, HRV, RUS MNE, SRB, UKR TJK, UZB
Policy and institutional foundation
Multistakeholder skills task force
National strategy on adult
learning
Coordination mechanisms
Autonomy
Promote autonomy of training
institutions
Accountability
National qualifications framework
Regulation and certification
regime
Certification and quality control
agency
Strategic financing
CVET—firms
Training tax deductions for firms
Training grant schemes based on
payroll levy
Grant schemes
SME support schemes, with training
focus
CVET—individuals
Worker tax deductions for
training
Subsidies or vouchers for
individuals
Training leave
Individual learning accounts
Loan schemes for individuals
Retraining and education for the
nonemployed
Job training for the unemployed
Second-chance education
programs
Source: Author’s assessment.
Note: Color key: priority, consider, not a priority. See “Acronyms” for country abbreviations and groupings.
populations and aspire to develop more advanced economies. In the case

of Turkey, the EU accession process will also focus on the Europe 2020
Strategy and thus, lifelong learning. In addition to developing effective
policy coordination mechanisms and strategies for adult education and
training, both countries should consider introducing regulatory systems
that use accountability and financing schemes to promote more labor
market-relevant adult learning programs.
Less Advanced Economies in Demographic Decline

The principal priority of less advanced economies in the region experi-
encing a demographic decline (i.e., many of those in South Eastern
Europe and the middle-income CIS countries) is to introduce a strategic
framework for adult learning and create the tools needed to implement
this strategy (e.g., coordination mechanisms, plus initial steps toward
regulation). Their demographic challenge and less advanced level of eco-
nomic development make adult learning critical for these countries.
However, in view of resource constraints, the competing claims of other
parts of the education system, and limited administrative and policy plan-
ning capacities, these countries should initially focus more on introducing
a strategy for adult education, and less on financing and regulation. At the
same time, this group can benefit from the experience of the EU10 (e.g.,
the Czech Republic) to learn how to foster competitive, private sector-
driven adult education and training markets.
Less Advanced Economies with Growing Populations

For the less advanced economies in the region that are not facing a demo-
graphic decline (i.e., those of low-income CIS countries and Albania), it
may be more productive to limit efforts to establish strategic frameworks
and coordination mechanisms for adult education and training. It is sug-
gested that this group of countries refrain from creating and piloting
financing schemes or implementing regulations until other educational
challenges, including the achievement of high secondary completion
rates, are adequately addressed. The more limited scope of action pro-
posed for this group is a specific reflection of the fact that (1) low-income
CIS countries face significant impediments to improving learning out-
comes in other education sectors, which will absorb much of their fiscal
and policy-planning capacity and (2) the relevance of adult education and
training to their economies is presently more limited, due to their differ-
ent demographic outlook and the greater distance of their economies
from advanced technology compared to the EU10 and middle-income

CIS countries.
Summary
Lifelong learning and adult education and training are growing in
importance across the ECA region. Increasing demand for skilled labor,
accelerating technological change, and significant demographic decline
are all contributing factors. Although building adult learning systems is
critically important, its overall priority in individual ECA countries
depends on economic and demographic conditions and their most
pressing education challenges.
Successful adult education and training systems are based on a high
degree of policy coordination and partnership between government
agencies and the private sector, with a strong voice for the demand
side—businesses and individuals—in determining training content. Once
a solid policy foundation has been created, governments can “steer” adult
education and training through tracking data on program quality
(derived from both routine monitoring and evaluation and impact stud-
ies), promoting autonomy and accountability, and ensuring efficient
government financing.
Across the board, private investment lies at the heart of adult educa-
tion and training, though governments will need to provide a certain
amount of financing to overcome market failures. In the case of CVET
programs for the employed, governments will mainly need to create an
enabling environment for private investment through appropriate regu-
lation and the provision of information. In addition, various financial
incentives for more private investment can be targeted to both firms and
workers. For the unemployed, retraining and education programs are an
important part of active labor market policies in most ECA countries; in
this case, governments can use their role as purchasers of retraining ser-
vices to shape the overall adult training market.
Notes
1. For a discussion of regional variation in SEE countries, see Gunny and Viertel
(2007). See also OECD (2005d) and Hungary, Ministry of Education and
Culture (2006).
2. The program also supports SMEs with training needs analysis and provides
opportunities for group training with other SMEs.
3. Despite its evident effectiveness, Turkey’s new Law on Labor Force Training,
adopted in 2008, abolished the placement incentive scheme.
4. Bulgaria, Romania, and Belarus lie close together in terms of GDP per capita
and the scale of their demographic decline. However, Bulgaria and Romania
should consider adult education and training more of a priority than should
Belarus, given the former countries’ membership in the EU, participation in
the Lisbon Agenda, and access to financing for lifelong learning from the
ESF.
CHAPTER 7
Extended Summary: The Path

for Education Reforms in the
ECA Region
The countries of Europe and Central Asia are coming out of a deep eco-
nomic crisis—the worst since the economic transition-induced recession
of the early 1990s and the deepest recession among all developing
regions. Post-crisis conditions look very different from those of preceding
years. Previously, growth in the region was fueled by cheap borrowing
from abroad, growing exports of goods and services (including labor ser-
vices), and, in the case of natural resource exporters, rising global demand.
The world of 2011 is a changed one. Borrowing is significantly more
expensive and export growth is restrained by potentially slower growth
in destination countries. Restoring and sustaining growth in this context
requires boosting competitiveness and increasing labor productivity—the
very reforms that were postponed in the boom years. These reforms are
all the more important given a shrinking working-age population in many
countries of the region.
The boom years also exposed significant bottlenecks to growth in the
ECA region, particularly with respect to the skills of the labor force
(Mitra, Selowsky, and Zalduendo 2010). Paradoxically for a region with
relatively high educational attainment, a shortage of worker skills
emerged as one of the two most important constraints to firm expansion
(the other being the tax regime). This finding makes it crucial for
193
countries to undertake reforms to reduce skills shortages as they plan for

recovery. Education systems play a very important role in creating the
right skills.
This book investigated two primary questions: Are education systems
in the ECA region able to inculcate the right skills in their graduates and
what can they do to improve on their performance? In answering these
questions, a fundamental problem emerged: data exist on the number of
students who graduate (i.e., how many diplomas are issued) in ECA
countries, but internationally comparable information on whether gradu-
ates of upper secondary and tertiary institutions (from which the bulk of
ECA graduates now enter the labor market) have the right skills and
competencies for the job market is not available.
A range of available data sources was examined to argue that there is
significantly more that education systems can do to improve the quality
and relevance of education and thereby enhance the skills of their stu-
dent populations. Although ministries of education are constrained in a
number of ways from effectively managing their education and training
sectors for results—that is, with an eye to student skills and
competencies—greater efforts should be made to reduce and eventually
eliminate the three most important and interrelated impediments to
educational success: the lack of systematic data on key skills-related per-
formance issues, the legacy of central planning, and the inefficient use of
resources.
Education and training systems in the region have simply been slow in
making the transition from teaching the basics (i.e., factual content, often
via rote learning) to inculcating the kind of higher-order skills increasingly
required in a modern economy. The ECA countries should therefore view
the reform of their education systems—at all levels and for all ages—to
be critical to their future economic success.
The following is an extended summary of the findings and policy rec-
ommendations of this study.
The Skills Challenge in the ECA Region

The ECA region had a well-regarded education system prior to the end of
central planning. Not only did countries in the region succeed in providing
mass basic education, they had a number of highly regarded universities
and research institutes, particularly in the basic sciences and mathematics.
While the intervening years have taken some of the shine off this reputa-
tion, these countries continue to have strong achievements. Notably,
enrollments are high at all levels of education. Primary enrollment rates
Extended Summary: The Path for Education Reforms in the ECA Region 195
are above 90 percent and primary completion rates are also very high
throughout the region. Although secondary enrollment rates in many
countries suffered a decline in the 1990s, they have since climbed and are
now at or above what might be expected in all ECA countries, given their
respective income levels.
The socialist legacy is particularly visible in low-income countries in
the region, which have the highest secondary enrollments in the world for
their income level. Tertiary enrollments, which have grown rapidly in the
past two decades, are also high relative to income levels, with the few
exceptions of certain low-income CIS countries, such as Azerbaijan and
Uzbekistan.
In addition to attainment, ECA countries also do well in delivering a
relatively good-quality education, particularly in the lower grades. In fact,
many younger students in the region outperform their Western European
students in reading at the fourth-grade level, scoring above the scale aver-
age on the Progress in International Reading Literacy Study (PIRLS) in
both 2001 and 2006.1 These results are not uniform across ECA, how-
ever. Countries that have not joined the European Union (EU), for
example, scored below the average, with the marked exception of the
Russian Federation, which scored well above the average and demon-
strated the greatest improvement between 2001 and 2006 (see figure
7.1). The commendable performance in reading at the elementary level
indicates that schools in ECA countries are very capable of providing a
quality education at the early stages.
Many ECA countries, though certainly not all, continue to do well on
international reading assessments such as PISA and TIMSS at or near the
end of lower secondary education, but only relative to the performance
of other countries at the same income level. Figure 7.2 provides a scatter
plot of 2008 per capita GDP against the PISA 2009 reading scores of
countries participating in PISA (with ECA countries highlighted by their
letter acronyms). The figure shows a positive relationship between the
two values, suggesting that 15-year-olds in countries with higher national
incomes tend to perform better on the reading component of the PISA.
Significantly, the ECA region seems divided. Many of the new EU mem-
ber states (including Poland, but also Turkey and Serbia) lie above the
regression line, demonstrating that their students apparently read better
than would be predicted on the basis of GDP per capita alone. Others lie
below the regression line, including Romania, Montenegro, Bulgaria,
Azerbaijan, and the Kyrgyz Republic. Students in these countries scored
lower on the reading component of the PISA than their GDP per capita
would have predicted.
196
Figure 7.1 Assessing a Skill—Reading Performance in the Fourth Grade: PIRLS Performance of ECA Countries, 2001 and 2006
600
550
mean performance
500
450
400
350
300
Macedonia, FYR
Bulgaria
Czech Republic
Hungary
Latvia
Lithuania
Poland
Romania
Slovak Republic
Slovenia
Georgia
Moldova
Russian Federation
Turkey
Balkans EU11 LI CIS MI CIS Turkey
2001 2006 PIRLS scale average
Sources: World Bank staff calculations based on PIRLS test score data downloaded from EdStats Database.
Note: LI = low-income, MI = middle income.
Figure 7.2 Analysis of Reading Competency of 15-Year-Old Students on the

PISA 2009
550
POL EST
500 HUN
LVA SVN
TUR HRV CZE
LTU SVK
450 RUS
SRB
BGR
ROM
MNE
400 KAZ
ALB
AZE
350
KGZ
300
0 20,000 40,000 60,000 80,000 100,000
Sources: PISA 2009 Database and World Bank staff calculations based on World Development Indicators Database.
Note: The figure shows a regression line representing countries’ predicted PISA reading scores based solely on
GDP per capita, compared to the OECD mean reading score (horizontal line) on the assessment and GDP per
capita in 2008. See “Abbreviations” for a key to country abbreviations.
In spite of these positive achievements of ECA education systems, the

EBRD-World Bank Business Environment and Enterprise Performance
Surveys (BEEPS) show that ECA firms’ perception of skills constraints
changed dramatically around 2005.2 By 2008, skilled labor shortages
had become one of the most commonly reported constraints to growth
in the BEEPS survey across all countries in the region, second only to tax
rates (see figure 7.3). On average, 30 percent of firms considered educa-
tion and skills to be a major or severe constraint in 2008. The highest
proportion of firms reporting constraints were found among the middle-
income CIS countries, where upwards of 40 percent of firms were dis-
satisfied with the availability of skilled workers. A smaller proportion of
firms in the Western Balkans reported similar levels of dissatisfaction,
with significant variability in the dissatisfaction levels of firms in the
EU10+1 countries and Turkey (BEEPS dataset 2008).
Why Are Skills an Emerging Problem if Education

Despite generally high average enrollment and attainment rates, as well as
respectable quality education for their income levels, especially at lower
levels of education (and even at lower secondary, where two-thirds of
participating countries “punch” at or above their weight), skills shortages
Figure 7.3 Distribution of Firms in ECA Region That Consider Worker Skills a
“Major” or “Very Severe” Constraint, 2008
9
mean = 30.2
8
TJK
7
UZB
number of countries
6
AZE ARM CZE
5
BIH GEO EST
4
KOS KGZ HRV UKR
3
MKD ALB LVA MDA
2
1
0
%
%
%
10
20
40
50
60
70
80
90
30
x<
x<
x<
x<
x<
x<
x<
x<
x<
<
<
<
<
<
<
<
<
%
%
%
%
60
80
10
20
30
40
50
70
Sources: BEEPS 2008; Mitra, Selowsky, and Zalduendo 2010.

Note: LI = low-income; MI = middle-income; x = % of firms (in respective countries) that consider education as
an obstracle. See “Acronyms” for a key to country abbreviations.
were constraining growth in ECA countries before the onset of the cur-
rent downturn. This is a puzzle that this book could not fully resolve,
mainly because crucial information is missing. International assessments
provide information about student competencies up to the age of 15 (usu-
ally the end of lower secondary education), but beyond that age, no infor-
mation exists on student competencies. Nor does information exist on the
quality of upper secondary or tertiary education or the relevance of educa-
tion at these levels. Moreover, no information exists on students’ noncog-
nitive (or socioemotional—“soft”—skills) (see World Bank 2011b for an
example of testing noncognitive skills in Peru). For policy makers to better
understand the role of education in contributing to skills bottlenecks—and
how to address it—this informational gap needs to be addressed.
Although the education offered in ECA countries at the end of lower
secondary is respectable relative to levels of income, it does not appear to
be adequate (or of the right relevance) to meet the demand for skills in the
region. In addition to higher levels of education, firms are demanding
higher-order skills that higher education alone does not guarantee, including
cognitive skills such problem solving, together with behavioral skills such
as teamwork (see box 7.1). Such skills are increasingly a part of the land-
scape of work in modern economies. Figure 7.4, which illustrates findings
from the European Work Conditions Survey, demonstrates how the cogni-
tive demands of work have risen relative to more routine tasks.3
The need for higher-order skills is not, moreover, confined to a few
high-technology sectors, even construction—which is not normally viewed
as a highly skilled industry—requires the ability to learn on the job, work
independently, and solve complex problems. Nor are higher-order skills a
requirement only of upper-middle income countries. Surveys of employ-
ers in lower-middle income countries, such as Kazakhstan and the former
Yugoslav Republic of Macedonia, point in the same direction: the skills
that are lacking include communication, thinking skills, problem-solving
skills, and the ability to work independently (Ivaschenko 2008, World
Box 7.1
Higher-Order Skills for the World of Work in the 21st Century

It is now recognized that beyond formal qualifications, what is needed to function
effectively in 21st century economies is the ability to use knowledge and other
personal and social skills to complete tasks and solve problems as they arise. This
broad group of abilities is generally called “skills” or “generic skills” because they are
not specific to a job and are transferrable across employers. While there is no
international consensus on key generic skills, most experts would agree that they
include both basic and higher-order skills. Basic skills include reading, writing,
computation, and the ability to use technology. Higher-order skills can be cogni-
tive (e.g., learning to learn, problem solving, and creative thinking) or behavioral
(e.g., communication, negotiation, and teamwork). The OECD list of key “compe-
tences” includes, in addition, the ability to act autonomously—that is, to perform
and conduct one’s own life plans—a skill that is particularly important in the mod-
ern world, where stable lifelong employment is less common. At present, stan-
dardized tests of secondary school students measure basic skills and some
higher-order cognitive skills, but not behavioral skills. Recent research in Peru pro-
vides an example of how to incorporate the testing of behavioral skills into regular
household surveys (see World Bank 2011b).
Sources: Authors, based on OECD 1999a and materials available on the following websites: State
Secretariat for Economic Affairs of Switzerland (www.seco.admin.ch), Australian National Centre for
Vocational Education Research (www.ncver.edu.au), American Society for Training and Development
(www.astd.org); all URLs accessed September 2010.
Figure 7.4 Worker Responses to Survey on Working Conditions in Europe, 2005
Does your main job involve: (% of respondents answering "yes")

90
80
70
60
percent
50
40
30
20
10
0
le en
th new
ta ng
ks
w ng
as
ob se
ex ki
of gi
s
gs
s
k
m
sk
or
st
pl rta
ds han
pr ore
ng
in
ou
m e
ni
f
co und
ho c
un
on
ar
et or
le
ot
g
m ng
in
on
lv
si
m
oo
so
g
ch
in
do
EU15 member states (i.e., non-ECA) new EU member states
Source: Authors’ calculations based on data obtained from Eurofound 2007.
Bank 2010b). These findings mirror longer-term trends observed in other

parts of the world (see chapter 1).
Many Students are Failing and Problem-Solving Skills are Weak

Two problems related to quality seem particularly acute in the ECA region:
too many students are failing and education systems have difficulties
imparting problem-solving skills. A closer look at ECA countries’ perform-
ance in international assessments show a large number of underperformers:
15-year-olds with such weak math and literacy skills that their ability to
succeed in today’s workplace is highly questionable (see box 7.2). As work
becomes more demanding even for this group, their poor skills will become
more apparent. Although ECA countries perform reasonably well on inter-
national assessments that measure students at the primary level (grade 4),
their performance is weaker on assessments of students around the end of
lower secondary education (grade 8), suggesting that their education sys-
tems are adept at imparting basic skills, but have problems imparting
problem-solving skills—the very skills that firms increasingly seek.
Another reason why firms are increasingly complaining may be that
the quality of education in the ECA region has not shown consistent
improvement. In many countries, the quality of education appears to be
Box 7.2
A Large Proportion of Students Are Failing

Despite PISA scores that are largely in keeping with countries at the same income
level, a large proportion of students in many ECA countries still achieved only a
basic level of competency (Level 1) or less on the reading part of the PISA 2009
(see table B7.2). This level means that a student can complete a basic reading task,
such as locating a piece of information or identifying the main theme in a text, but
cannot necessarily use this information in problem solving. The factors underlying
this poor performance vary. In Turkey, a nontransition country, a very selective
academic system leaves many graduates of vocational and general high schools
with poor basic skills. In Azerbaijan, poor reading competencies co-exist with
strong competencies in mathematics, largely because the transition from a Cyril-
lic to Latin alphabet in the post-independence period left a generation of stu-
dents and their parents without adequate reading materials to build
reading skills.
Table B7.2 Proportion of 15-Year-Old Students in ECA Who Achieved Only

Basic Reading Competency on PISA 2009
Students scoring at or
below Level 1 Country
< 10 %
10–20 % Estonia, Poland, Latvia, Hungary, OECD average
20–30 % Croatia, Czech Republic, Lithuania, Slovak Republic, Slovenia,
Russian Federation, Turkey,
30–40 % Serbia
40–50 % Bulgaria, Montenegro, Romania
50–60 % Albania, Kazakhstan
> 60 % Azerbaijan, Kyrgyz Republic
Source: PISA 2009 Database.
getting worse, not better. Rather than narrowing the gap between their
scores and OECD mean scores, many ECA countries that participate in
international learning assessments at this level—which are likely the
better-managed countries—either seem stuck or have regressed in terms
of their scores. It is probable, moreover, that the situation is even worse
in ECA countries that do not participate in efforts to measure and com-
pare the academic performance of their students. In fairness, the latest
round of PISA (2009) showed some improvements for 11 ECA countries,
compared to their 2006 performance (including big increases in Bulgaria,

Romania, Serbia, the Kyrgyz Republic, and Turkey). However, seven
ECA countries experienced a deterioration compared to their 2006
performance.
The Quality and Relevance of Upper Secondary and Tertiary

Education Is of Concern
An additional explanation of the skills gap perceived by employers may
be that the quality of upper secondary and tertiary education in the
region is not keeping up with changing skill demands, although the lack
of data makes it hard to confirm this suspicion. With educational quality
failing to show consistent improvement at the lower secondary level, this
weakness is probably mirrored at the upper secondary level (although
with a delay). Moreover, enrollments at the tertiary level used to be
tightly controlled and reserved for the few, best performers. In the past
two decades, however, enrollments have doubled, tripled, or quadrupled
in an environment with the expansion of new programs, institutions and
types of students. This growth has taken place without the benefit of
quality assurance mechanisms and without the kind of information stu-
dents and parents need to make informed choices. There is also evidence
of widespread unethical behavior, such as purchasing admission, grades,
and even degrees (see figure 7.5). Educational quality at the tertiary level
is thus unlikely to have improved. Weaknesses in the quality of upper
secondary and tertiary education is particularly problematic from a skills
perspective because, most of ECA’s new labor market entrants now enter
the labor market after having completed either an upper secondary or a
tertiary degree (see chapter 3, figure 3.1).
Finally, graduates of upper secondary and tertiary education may also
be graduating with the wrong set of skills. During the early years of the
transition, the vocational school system—which once produced more
than half of all secondary graduates in most ECA countries—rapidly
declined. Since then, students have left vocational schools in favor of
general secondary education and the prospect of pursuing a tertiary
degree. However, ECA firms still value technical and vocational skills,
judging by the better employment prospects of graduates with technical
skills (see chapter 1). Thus, the pendulum may have swung too far in one
direction, with the vocational sector having retrenched too far.
Nevertheless, as the region’s vocational schools remain unable to produce
graduates with appropriate skills, it is likely too soon to be promoting a
return to this sector.
Figure 7.5 Students Report Unofficial Payments Are “Usually” or “Always” Needed
in Public Technical Colleges and Universities, 2006
15
% of respondents
10
0
at er
ro rn
IS
CI
IC
b
Eu ste
es
pe
em
LI
M
Ea
m
st
h
EU
ut
So
usually always
Source: Raw dataset from EBRD-World Bank 2006.

Note: LI = low-income, MI = middle-income. The graph reflects the answers from Question 3.13 of the Life in
Transition Survey in ECA countries, which can be downloaded at http://www.ebrd.com/downloads/research/
economics/litsques.pdf (accessed January 2011). The exact question asked was: “In your opinion, how often is it
necessary for people like you to have to make unofficial payments/gifts in these situations?” One of the response
options was: “Receive public education (university, college).” Thus, it is unclear whether the response refers to
grades or admission, or simply to regular payments.
Adult Learning Remains a Blind Spot

The skills deficit relates not only to the uneven quality and uncertain
relevance of formal education. It also involves too little progress in pro-
viding workers with options for additional training. Since continuous
adult education and training have been shown to foster employment and
greater productivity, the development of this sector should be central to
the region’s economic growth strategy, particularly in more advanced
ECA countries and those that are facing a significant demographic
decline. Yet many of these countries have only started to plan for the
development of this sector. In light of the growing skills shortage, adult
education and training in ECA countries can no longer be neglected.
Priority Areas for Action

As noted previously, three major impediments are inhibiting the creation
of flexible and responsive education systems in the ECA region. First,
these systems have been operating “in the dark” because they do not sys-
tematically collect data on the learning and employment outcomes of all
students. Second, the legacy of central planning has kept the governance
and management of school systems highly centralized, with central policy
makers deeply involved in operational details. While virtually all OECD
countries have embraced performance-oriented management in educa-
tion since the 1980s (see OECD 2005c)—albeit at different speeds—
most ECA countries continue to use management practices that focus on
compliance with detailed regulations and financing schemes based on
inputs, not outputs. This means that most local education authorities and
school principals in the ECA region lack the autonomy and authority to
make crucial management decisions for their own institutions, including
how much and what type of vocational content students can choose from
and how many teachers to employ. As a result, the education and training
system is inflexible and does not respond either to labor market needs or
to changes in student numbers.
Third, financial resources are ineffectually used in the region, particu-
larly in the pre-university sector, where few countries have adjusted teacher
staffing levels in response to falling student numbers over the past 20 years.
As a result, student-teacher ratios (STRs) have fallen sharply as per student
costs have risen—more so than in any other region in the world (see figure
7.6). This implies that scarce resources are misused by paying meager sala-
ries to too many staff members and heating half-empty buildings.
With few exceptions (notably Turkey), class sizes in ECA are also
smaller than those in other parts of the world. For instance, the average
class size at the primary level in Poland, Serbia, and Azerbaijan is 20, 14,
and 12, respectively, compared to 23 in France, 26 in the United Kingdom,
and 31 in Chile and the Republic of Korea (see figure 7.7). Although the
ECA region also has a large number of small schools, relatively small class
sizes are prevalent even in the largest schools, with the possible exception
of Armenia.
Small class sizes and lower STRs in the ECA region have not, how-
ever, improved educational quality—as the cost per student has
increased, student learning outcomes have not improved. Instead, these
conditions have led to poor teacher remuneration—especially at the
entry level, but also relative to other professions—and ultimately, to the
demoralization of the teaching workforce. This has discouraged talented
young students from becoming teachers, while simultaneously encourag-
ing those already in the profession to stay on as long as they can to
benefit from seniority-related pay. As a consequence, the profession as a
Figure 7.6 Primary School Student-Teacher Ratios in ECA Compared to Other

Regions of the World, 1990–2008
29
27
pupil-teacher ratio (primary)
25
23
21
19
17
15
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
Sources: EdStats Database, plus authors’ linear interpolations for missing years.
(in both the World Bank’s EdStats Database and other international education databases) is that it is unclear
whether or not data for a particular country are reported on a full-time equivalent basis.
whole is failing to renew itself: few new recruits are entering the system
and the teaching cadre is aging (see figure 7.8). As the wage bill squeezes
out other spending, less resources are available for training and the training
received is often of questionable value. Given the importance of effec-
tive teaching, especially for inculcating higher-order cognitive skills,
ECA countries are thus poorly positioned to improve student learning
with their existing teaching workforce.4
To be fair, these impediments affect ECA countries differently and
vary accordingly to the level of education. However, no ECA country has
fully escaped the legacy of central planning.5 That system focused on
generating data on inputs—that is, it checked whether local actors were
in compliance with detailed norms for all inputs. Education ministries
remained in the dark, however, about the return on those inputs, that is,
whether students actually acquired skills and competencies.
In terms of how these impediments affect the different levels of educa-
tion, they are most clearly apparent at the pre-university level, which
absorbs two-thirds of total education funding in the ECA region. The
206
Figure 7.7 Comparison of Average Size of Primary School Classes Worldwide, Various Years
35
average class size (number of students)
30
25
20
15
10
0
Slovenia
Czech Republic
Estonia
Slovak Republic
Poland
Hungary
Bulgaria
Azerbaijan
Armenia
Russian Federation
Ukraine
Turkey
Serbia
Luxembourg
Greece
Spain
Finland
Germany
France
Ireland
United Kingdom
Mexico
United States
Brazil
Chile
Republic of Korea
OECD average
EU-11 LI CIS MI CIS Turkey Balkans Western Europe other
Sources: OECD Education At a Glance Database for OECD and partner countries; authors’ calculations based on Ministry of Education data (from education management information
systems) for Azerbaijan, Serbia, and Romania, and on Statistical Institute school-level data for Bulgaria.
Note: LI = low-income, MI = middle-income. Most recent data for each country (2005, 2006, 2007, or 2008).
Figure 7.8 Percentage of Students with a Teacher Over 50 Years Old in ECA
Countries, Selected Years
60
% of children with teachers
50
over 50 years old
40
30
20
10
0
Bulgaria
Hungary
Lithuania
Romania
Slovak Republic
Slovenia
Macedonia, FYR
Turkey
Russian
Federation
Armenia
EU10+1 SEE MI CIS LI CIS
countries
1995 1999 2003 2007
Sources: IEA (1995, 1999, 2003, and 2007): Mathematics and Science Teacher Background Data Almanacs.
same three impediments affect tertiary education in a slightly different

way, as this sector has already undergone significant reform over the past
20 years, during which time it has doubled or tripled in size in many
adult education sectors in the modern sense, these impediments cannot
be fairly characterized as problems of this sector. However, the policy
recommendations offered in this book do apply to adult education and
offer ideas and principles for how this subsector could be developed and
managed in the region (see chapter 6).
Managing Education Systems for Results

This book argues that in order to address the three constraints outlined
above, countries in the ECA region need to manage their education sys-
tems for results. First and foremost, they need to collect more information
on “performance,” that is, what are students learning, what jobs they are
finding, and what salaries graduates are earning. In addition, ECA coun-
tries need to follow the lead of OECD countries and replace input-
oriented management of the education sector with performance-oriented
management. This change implies devolving more autonomy to frontline

providers and putting in place accountability mechanisms—for instance,
performance contracts and performance-based budgeting—that empha-
size results and performance, not compliance with norms. With greater
availability of performance data and frontline providers empowered with
greater decision-making power, the preconditions will be in place for
education systems in the region to become both more responsive to labor
market needs and more efficient.
Focus Attention on Learning and Employment Outcomes

Better data alone does not lead to better-quality education. Partly due to
difficult demographic trends and partly due to history, ECA education
systems have wound themselves in a knot that will be very difficult to
untie without devoting greater attention to results. The legacy of central
planning has created a system where bureaucrats manage the sector, and
schools in particular, by writing detailed norms into legislation and then
micromanaging principals so that they comply with these norms.
Generations of managers have lived with this system for their entire lives.
When asked to improve education system performance, they therefore
reach for the tools that they have always used, either by revising norms
or ramping up inspections. This chapter argues that the solution is not to
rewrite norms or expand the inspectorate. Rather, the solution is to reach
for different tools. However, none of these “new” tools will work without
better data on performance, data that answers the question: What are
students learning and are graduates finding jobs?
The good news is that there is significant consensus on the key skills
and competencies that basic education is expected to deliver, as well as
standardized tests to measure them. Many countries in the region already
participate in international assessments, which are more useful when
complemented by national assessments. Although all school systems in
the region have national assessment systems, many of these systems are
in the early stages of development (see figure 7.9) and there is very little
evidence to date that ECA countries are using these assessments to
inform policy. Rather, national assessments continue to be seen as a certi-
fication mechanism, one that confirms a student has gained a mastery of
a predetermined curriculum, thus enabling a diploma to be issued or the
student to progress to upper secondary or tertiary education. Student
assessments are rarely seen as an opportunity to identify the strengths and
weaknesses of an education system.
When students graduate, ministries of education in the ECA region do
not systematically collect, analyze, and disseminate information on their
Figure 7.9 Status of Measuring and Using Data on Student Learning Outcomes in
Several years of
experience in
Several years of measuring,
Early piloting of participation in improve education
some participation
in international Bulgaria
Not started, very assessments
Hungary
early stages, no, or
Albania
assessments Romania
Croatia
Armenia Serbia
Czech Republic
Azerbaijan Slovenia
Estonia
Belarus
Bosnia and Georgia
Herzegovina Macedonia, FYR
Kazakhstan Moldova
Kosovo Montenegro
Kyrgyz Republic Poland
Tajikistan Russian Federation
Turkey Slovak Republic
Turkmenistan Ukraine
Uzbekistan
in PISA, PIRLS, and TIMSS since 1995; and inputs from World Bank country experts. The UNICEF data is drawn
from “Table 2.1: Status of Reforms of Assessment and Examination Systems, 2006,” regarding “Introduction of
other school exams or assessments (e.g., basic school)” and “Introduction of sample-based national assessment.”
In this table, UNICEF scores the progress of countries on a range of 0 to 4, with 0 representing “not planned or
started” and 4 representing “operational.” In addition, the figure uses World Bank staff compilations on the
number of international assessments in which each country has participated.
employment outcomes. Such information is, however, critical for policy

makers and higher education institutions because it helps them detect
which programs and fields of study are in high demand among employers.
Moreover, these data can help students make better choices about which
university and field of study to pursue. Again, some member countries of
the OECD (including Hungary, Italy, and the Netherlands), together with
Romania, provide examples in this area for others to emulate. Using the
example of the tracer study in the Netherlands (which has been in place
since 1989), table 7.1 provides an example of the types of information
such surveys can provide.
At the tertiary level, data on employment and learning outcomes are
also needed to shed light on how the sector is performing from a skills
perspective. Currently, data on tertiary outcomes (both in ECA and
around the world) focus on research outcomes, such as how many publi-
cations the faculty of a university is producing. Unfortunately, when this
is the only performance indicator available, it inadvertently receives more
attention than it should. Indeed, rankings of universities in the region
place significant weight on the number of publications that they produce.
However, from a skills perspective, the outcome that matters most is
whether students are graduating with the competencies in demand on
the job market and finding jobs. Currently, these outcomes are not being
measured, and, unfortunately, an international standardized learning
assessment for tertiary students is not expected to be launched by the
OECD until 2016 (see box 5.1). Thus, to gather the necessary data
now—when it is needed—ECA countries will need to begin with their
own, domestically developed approaches.
There are several complementary ways of providing more information
on the performance of the tertiary sector, some of which are more direct
than others. These include rankings or league tables, tracer studies, and
direct measurement of student competencies via standardized tests. One
way that ECA countries might begin to measure the learning outcomes
of tertiary students would be to introduce standardized testing within
fields of study where these outcomes have been clearly defined by the
National Qualification Frameworks which have been built in most coun-
tries (e.g., for graduates of nursing programs). Alternatively, ECA minis-
tries of education might define and measure a set of broad competencies
that all tertiary graduates are expected to have. The OECD’s DeSeCo
project (OECD 2005c), discussed in chapter 2, could provide a starting
point for the definition of such broad competencies.
Introduce Autonomy and Accountability Based on Results

Overcoming the legacy of central planning involves moving away from
detailed norms and instead holding actors accountable for performance
(or results). At the central level, this implies that policy makers in the
ECA region would relinquish certain duties and assume others so that
education systems may innovate and improve student learning outcomes.
The opportunity before central governments in these countries is to
Table 7.1 Information Collected from Tracer Study of Dutch University Graduates, 2007
Duration of Full-time Unlimited Managerial or High High job
job search employment term contract Monthly gross professional utilization of satisfaction
(in months) (%) (%) income (euros) (% ISCO 1 or 2) skills (%) (%)
Mean Mean Mean Mean Median Mean Mean Mean
Science and math 0.7 88.4 45.6 2499 2429 83.6 71.0 77.5
Medicine and health 0.7 79.0 41.0 2904 2783 82.6 78.9 81.7
Engineering 1.0 94.4 66.2 2772 2631 87.4 75.2 72.8
Economics 1.0 96.1 71.3 2954 2783 70.3 65.6 70.1
Law 1.2 92.7 57.5 2864 2732 87.8 66.0 70.2
Humanities and arts 1.2 59.1 43.0 2188 2226 66.0 50.4 61.3
Social sciences 1.3 60.6 45.4 2317 2350 72.0 63.9 65.4
Agriculture 1.5 86.7 45.6 2137 2328 84.2 73.2 71.2
Source: Table provided by the Research Centre for Education and the Labour Market (ROA), The Netherlands, 2008, at request of the authors.
211
move away from micromanaging schools and classrooms and focus

instead on setting goals, policies, and standards; defining responsibilities;
mobilizing financial resources; ensuring political consensus; targeting
poor and excluded students; and monitoring and evaluating service deliv-
ery and system quality.
authority and responsibility in the hands of the people most able to inno-
vate and improve the quality of education: local managers and education
authorities (see Osborne and Gaebler 1992). It also means holding these
actors accountable for improving the learning outcomes of the bulk of
their students. Aligning the incentives of these stakeholders with the stu-
dent learning outcomes desired by policy makers requires education
ministries to set overall performance goals; articulate who is responsible
and accountable to whom and for what; and ensure that these responsi-
bilities are agreed, accepted, and understood.
At the pre-university level, policy makers can expand the autonomy of
lower-level actors in the school system by granting them greater decision-
making power over school operations and budgets and relaxing norms on
class sizes. Greater autonomy is particularly needed in vocational educa-
tion and training, where programs and institutes need the ability to
expand or contract course offerings in response to student and employer
demand, not rigid governmental norms.
Simultaneously, policy makers will need to implement a range of
accountability mechanisms to hold schools accountable for results.
Potential mechanisms include the creation of school councils that involve
parents and local communities in school decision-making processes;
requiring schools to prepare school development plans that outline each
school’s strengths and weaknesses, together with an action plan for mak-
ing improvements; preparing “school scorecards” that include basic indi-
cators on a school’s performance; as well sanctioning low-performing
schools and rewarding high-performing schools (see chapter 4 and box
7.3). Here, however, it is important to define a high-performing school as
one that delivers improved learning outcomes to all types of students,
including those from low-income and minority households.
In general, the weaker a country’s national assessment system, the
more it will need to rely on more qualitative measures to assess school
performance. It is important to stress here that utilizing information on
student learning outcomes to strengthen accountability does not neces-
sarily imply “paying for performance” or conducting tests that have very
high stakes for principals, local authorities, and teachers. If not carefully
Box 7.3
Options for Making Schools More Accountable

for Learning Outcomes
Administer national student assessments at key stages of the education cycle
(e.g., after 4th, 7th, and 12th grades) and include such data in school “scorecards.”
Require schools to prepare school development plans that outline each
school’s strengths and weaknesses, together with a vision and action plan for
making improvements. Support implementation of these plans and monitor their
outcomes.
Reward schools that show improvement in student performance.
Prepare school “scorecards” for all schools that include information that
enables parents and students to get a sense of their performance. Make such
scorecards available on the website of the ministry of education, as well as dis-
seminated to local authorities and schools.
Agree on criteria that objectively identify schools in need of improvement and
decide what actions should be taken to support such schools.
Sanction schools that show low or no improvement in student performance.
Require all school principals to be trained and licensed and be held more
accountable for learning outcomes of the school’s pupils.
designed, such an approach could create exactly the wrong incentives for
improving learning. Rather, policy makers should use performance data
to identify both schools that need greater support and schools that offer
successful models for innovation.
unless policy makers in the region address the demoralization of the
teaching force, an effort that will involve resolution of serious overstaffing
and improving the pay, professional development, and work conditions of
teachers.
Most countries in the region have shied away from adjusting teaching
staff levels in response to declining student numbers, even in contexts
where per student financing provides powerful incentives to reduce
teacher numbers. Addressing this inefficiency will free up funds for
more classroom learning materials. Judicious use of retirement incen-
tives for older teachers and stronger redundancy or retraining packages
for departing teachers or both may make departures more acceptable.

Countries now in the process of reducing teacher numbers (e.g., Bulgaria,
Latvia, and Romania) may provide valuable lessons in this area.
At the tertiary level, policy makers in most of the ECA region have
already granted substantial autonomy to institutions of higher learning,
which now manage their own budgets and make important managerial
decisions about their respective institutions. This autonomy has helped
create some (but not all) of the conditions needed to improve the quality
of course offerings and pedagogy with the goal of preparing students for
the world of work. Policy makers have also begun to use financing incen-
tives to “steer” these institutions towards education of greater quality and
relevance to employer needs. In fact, considerably more countries in the
region have adopted per student financing in the tertiary sector than in
the pre-university sector. In terms of accountability, all countries in the
region (with the exception of four central Asian countries and Kosovo)
have become signatories to the Bologna Process, which aims to make
academic degree and quality assurance standards more comparable and
compatible across Europe.
Today, the challenge of tertiary education systems in ECA is to
wield the instruments of autonomy, accountability, and performance-
based financing more effectively with the goal of rapidly improving
the quality and relevance of higher education. As noted earlier, most
of new EU member states and, to a lesser extent, countries in SEE, the
Russian Federation, and Ukraine, have devolved a fair degree of auton-
omy to tertiary institutions and have taken initial steps toward flexible
financing. However, this group has lagged behind in establishing effec-
tive accountability frameworks. Certain countries in SEE and low-
income countries in the CIS have retained more centralized control
over tertiary education systems and would benefit from moving ahead
to introducing greater autonomy, flexible financing, and accountability
mechanisms.
Indeed, it makes little sense to delay in enhancing accountability in the
tertiary sector. In the first place, greater autonomy and flexible financing
will be ineffective without mechanisms that hold education providers
accountable for the quality of the education that they provide. In the
second place, it takes considerable time to develop and implement these
mechanisms effectively. Many countries that have moved ahead on flex-
ible financing and autonomy without strengthening accountability have
later had to claw back control over tertiary institutions—a process that
can be politically difficult.
Strengthening accountability at a minimum means safeguarding basic

integrity in higher education. On the financial side, this means preventing
embezzlement, fraud in public tenders, collusion, and the hiring of
unqualified family members, among other practices. On the academic
side, this means preventing examination fraud, unethical faculty behav-
ior, noncompliance with admission standards, research fraud and plagia-
rism, and fraud in the quality assurance process (Salmi 2009). With the
exception of a few new EU member states, evidence suggests that higher
education institutions across the region continue to struggle with these
basic integrity issues.
While integrity problems are widespread, that does not mean that
policy makers need to resolve these problems first, before they make
progress on accountability in other areas. The great temptation in the
region—given its tradition of central planning and central control of
universities—is to fix integrity problems through even stronger central-
ized control and management (e.g., requiring that every hiring and
procurement decision be reviewed by the central education ministry).
Rather, policy makers need to embrace solutions that show promise in
terms of addressing basic integrity and strengthening accountability.
The good news is that most of the instruments that help improve basic
academic and fiscal integrity also strengthen accountability for desired
outcomes and the efficient use of resources.
Looking at the different options illustrated in table 7.2, it is clear that
strengthening accountability in the tertiary sector can be a long and
Table 7.2 Tools for Strengthening Academic and Fiscal Integrity in University-Level
Institutions
Tools that strengthen academic integrity. . . Tools that strengthen fiscal integrity…
• Licensing • Financial audits
. . . and educational quality . . . and efficiency of spending
• Accreditation/academic audits/evaluations • Public disclosure laws
• Public disclosure laws • Strategic budget plans
• Fostering outsiders to review academic • Performance contracts based on
integrity (e.g., Romania) performance indicators
• Performance contracts based on perfor- • Student loans/scholarships/vouchers
mance indicators
• Creation and dissemination of rankings/ • Embedding incentives into allocated
benchmarks resources (e.g., tying funding to the
number of students enrolled or graduated,
rewarding good performers, etc.)
Source: Authors’ review of available instruments, as presented in Salmi (2009).
difficult process. Building an effective accreditation agency or a national

quality assurance system, for example, is a long-term proposition. Building
the capacity of a national quality assurance agency to fully meet its man-
date is a challenge for all countries in the ECA region. Many smaller
countries, for example, simply do not have the in-country capacity to
conduct effective national quality assurance. Here, coordination across
countries and closer integration with external quality assurance agencies
may be important to support the development of national capacity.
In light of these challenges, countries would be well-advised to fol-
low a two-pronged process to strengthen accountability: first, build an
effective quality assurance system that ensures tertiary institutions
have the minimum levels of inputs and processes needed to provide a
quality education; and second, emphasize greater transparency and
information dissemination, so as to strengthen student choice (e.g., by
developing rankings). These two approaches are complementary to
one another.
As noted earlier, most ECA countries are signatories to the Bologna
Process. Proper implementation of its 2005 Standards and Guidelines for
Quality Assurance would go a long way to putting in place the building
blocks of an adequate quality assurance system. A large number of coun-
tries in the region have not, however, begun the process of strengthening
external quality assurance. In fact, a number have treated the strengthen-
ing of quality assurance within higher education institutions (so-called
internal quality assurance) in a rather cavalier fashion. For the five ECA
countries that are not yet signatories, aligning themselves with the
Bologna Framework would be beneficial.
Improve the Efficiency of Resource Use through

Performance-Based Financing
Managing school systems for performance means moving away from
inflexible line-item budgeting towards greater use of delegated budgets,
with incentives for maintaining enrollment and attendance (i.e., per stu-
dent financing—a path on which the majority of ECA countries have
already embarked). More flexible, smarter financing in the form of block
grants (i.e., contract- or performance-based) can also provide funding to
local education in return for meeting agreed learning outcomes. Not only
does this type of financing give local education authorities much-needed
flexibility that input-based budgets do not permit, it keeps them focused
on student results.
At the pre-university level, countries in the ECA region are at different

stages of introducing per student financing (see figure 7.10); the experi-
ence of those that have gone farther can provide lessons for countries that
have yet to start on this path. Whether per student financing actually
results in greater fiscal efficiency and an increased focus on student learn-
ing outcomes depends on the financing formula used and the broader
context in which it is implemented. If, for example, a financing formula
Figure 7.10 Progress Towards Results-based Education Financing in the ECA

Region, 2009
per student
financing widely
introduced
piloting of per
student financing
Armenia (2005)
Bulgaria (2008)
per student Kyrgyz Republic
financing under Czech Republic
(2006)
discussion (1992)
Moldova (2010) Estonia (2001)
Russian
Georgia (2007)
Azerbaijan Federation (1998)
not started or Hungary (1990)
Latvia Tajikistan (2005)
very early stages Kosovo (2002)
Macedonia, FYR Uzbekistan (2008)
in discussions Lithuania (2001)
Moldova
Poland (2000)
Serbia
Albania Romania (2010)
Slovenia
Belarus Slovak Republic
Bosnia and (2004)
Herzegovina
Croatia
Kazakhstan
Montenegro
Turkey
Ukraine

Note: The “result” referred to in this table is a student enrolled. As discussed in chapter 6, a more desirable per
student financing scheme would be to finance a student who is graduating, or better yet, a student who is
graduating with a desired level of competencies. However, even the most advanced ECA countries (in terms of
moving to school financing based on results) still base financing on the number of students enrolled (an inter-
mediate result, at best). This report therefore intentionally distinguishes between the inputs of teachers, classes,
and finances, and between students “enrolled,” “graduating,” “graduating with a desired level of competencies,” or
“graduating and finding a job” as results (or outputs).
mirrors current unit costs of different localities, it may simply perpetuate

an inefficient delivery model. In the Slovak Republic, per student recur-
rent costs are 100 percent higher for upper secondary vocational educa-
tion and sports education schools than for gymnasia (upper secondary
academic schools). These differences in unit costs, however, are largely
caused by differences in class sizes and teaching loads, factors that should
not necessarily be encouraged to continue (World Bank 2009b). By con-
trast, countries that have devised simple, transparent formulas which
reflect the real costs of schooling have created strong incentives for school
consolidation, increased class sizes, and a rebalancing of educational
spending from wage to nonwage components (see the discussion of
Bulgaria in chapter 4).
At the tertiary level, several ECA countries (for example, Poland and
Romania) have already adopted per student financing, shifting their focus
away from inputs and toward outputs. The remainder of countries in the
region will need to introduce performance-based financing to improve
the efficiency of spending in this sector.
Beyond per student financing, there is still relatively little use of “per-
formance contracts” and “performance-based budgeting” in the region
(see OECD 2007c). Such contracts (and budgeting arrangements) could
be put in place between the ministry of education and universities, or
simply between the ministry of education and its key agencies (which
are charged with various roles in the management of the university sec-
tor) as discussed in chapter 5. Most ministries of education in OECD
countries have moved toward such arrangements to enable ministries to
focus on strategy and system monitoring rather than day-to-day opera-
tional details.
Another important move in terms of financing would be for ECA
countries to consider more cost sharing at the tertiary level. Indeed, if
ECA countries want to simultaneously raise participation rates and
increase the quality of tertiary education, more private resources will
also be needed—the cost pressures facing higher education in the
region are simply too great. Mobilizing private resources will help
strengthen results-oriented outcomes in the tertiary sector in two fun-
damental ways: (1) by increasing competition, as more nimble and
innovative private providers help bring innovations to public providers;
and (2) by increasing students’ connection to the education process.
When students and parents pay tuition fees—whether to a public or a
private provider—they are generally more demanding about the quality
and relevance of the education provided. There is no single ideal level
of funding for higher education and no single ideal mix of public and
private funding sources. Different countries will make different kinds
of trade-offs.
Build the Foundations of Adult Learning Systems

The reskilling of the work force in the ECA region is needed not only to
compete in the global economy, but also to address the lag effects of the
transition to market economies. Significant segments of the labor force—
often comprising less skilled and more disadvantaged workers—remain
inactive in many of these countries. Given the rapid demographic decline
occurring in many ECA countries, achieving and sustaining high rates of
growth suggests that labor force participation will have to increase and
each individual will have to become more productive—and stay so for a
longer period of time.
Unlike pre-university and university education, adult education sys-
tems in the modern sense are largely nonexistent in most ECA countries.
Building such systems will require shifting away from government-de-
fined programs towards a well-regulated market of private and public
providers that deliver training services to both working and unemployed
adults. Governments in the region will, however, continue to play a role
in education and training for the unemployed, often by contracting
private providers to deliver needed services. Policies that address market
failures in this sector are especially important for laying the groundwork
for effective adult education systems. In general, successful systems
require a high degree of coordination and partnership between govern-
ment agencies and the private sector, as well as giving the demand side of
training—that is, businesses and individuals—a strong voice in determin-
ing training policy. Once a solid adult learning sector is established, gov-
ernments can then “steer” it by monitoring data on program quality,
encouraging autonomy and accountability, and improving the efficiency
of government financing in the sector.
region that are facing a demographic decline. Their challenge is to ensure
that existing coordination mechanisms function well and that regulation
and financing are used to facilitate the emergence of a private sector-
oriented adult education and training market. The principal priority of
less advanced economies in the region experiencing a demographic
decline (i.e., many of those in South Eastern Europe and the middle-
income CIS countries) is to introduce a strategic policy framework for
adult learning and create the tools needed to implement this strategy
(e.g., coordination mechanisms, plus initial steps toward regulation). Less
advanced economies in the region that are not facing a demographic
decline (i.e., low-income CIS countries and Albania) may find it more
productive to limit efforts to establishing a strategic policy framework
and coordination mechanism for this sector. For many countries, partici-
pating in the OECD’s Programme for International Assessment of Adult
Competencies (PIAAC) would be an important first step in understand-
ing the current skills and competencies of their work forces.
Summary
Reforming education systems in the ECA region so that they deliver
higher-quality and more relevant education to the majority of students
will go a long way towards addressing the skills gap that many countries
in the region face. Although the recent economic downturn has dimin-
ished the demand for labor and skills. Skills shortages will likely reemerge
once growth picks up again. The process of reforming education will take
time; consequently, ECA countries should not wait to begin.
Many needed changes—assessment systems that measure what stu-
dents learn, changes in the size and remuneration of the teaching force,
and strong quality assurance systems—are not without cost. Given that
economic growth is expected to be lower and fiscal constraints tighter in
the near to medium term, ECA countries will have little choice but to
free up resources that are currently funding existing inefficiencies in their
education systems, particularly those associated with school infrastruc-
ture and staff levels designed for much larger student populations. More
resources will also need to be raised from outside the education system,
especially at the tertiary level. Putting off the policy reform directions
discussed in this book—whether for financial or political reasons—risks
the deterioration of a fundamental source of national competitiveness: a
skilled workforce. This is clearly not an acceptable option for any country
in the region.
For movement of education systems away from an emphasis on inputs
and processes to an emphasis on improving the learning outcomes of the
majority of students, much more measurement of student learning out-
comes is needed at all levels of education. This data must then be used in
the policy process to identify strengths and weaknesses and help formu-
late appropriate policy responses. As the role of central governments is
redefined to focus on strategic policy, system goals, and regulation, central

policy makers can articulate clear goals for their education systems and
use incentives to manage lower-level actors in these systems to meet
these goals. In schools, this means greater use of per student financing,
more autonomy, and greater accountability for learning outcomes. At the
level of higher education, this means strengthening national quality assur-
ance mechanisms in the short and long term, greater dissemination of
information (e.g., rankings) to inform student choice, greater perfor-
mance-based budgeting, and, where warranted, greater institutional
autonomy.
It is difficult to draw conclusions about which reform options are the
most relevant to the various subregions of Europe and Central Asia, as
there is enormous variation in their current conditions. What is clear is
that some member countries of the EU, together with certain countries
in South Eastern Europe and the Commonwealth of Independent
States, have not yet begun to systematically measure student learning.
What is also clear is that virtually all countries in the region are not yet
using assessment results to design education policy, a process that is
complicated by the fact that assessment data—where it exists—often
cannot be sufficiently disaggregated to identify problem schools and
groups of students. A similar situation holds for per student financing:
the greatest progress has been made in this area by the EU10, but prog-
ress has been highly variable in other subregions. It is also clear that all
countries in the region need improved accountability mechanisms in
the tertiary sector.
The economic crisis provided many countries in the region a reason to
initiate reforms, particularly those linked to the financing of education.
Under pressure to reduce spending, countries such as Romania, Poland,
Serbia, and Latvia introduced per student financing in their general edu-
cation sectors (covering primary and secondary education). This policy is
likely to result in greater efficiency and open the way for future spending
increases to focus on improvements that enhance the quality of educa-
tion. These countries will, we hope, establish a path that can be followed
by countries that have not yet focused on increasing the efficiency of
educational spending.
Increasing demand for skilled labor as growth picks up, accelerating
technological change, and significant demographic decline will soon make
modern adult education and training systems an imperative in many
countries in the region. Although building such systems is critically
important for reskilling adult workers and retaining workforce productiv-

ity, its overall priority in individual ECA countries will depend on their
specific economic and demographic conditions.
While the economic crisis offered opportunities for reform, it also
made certain reforms more difficult. Both Bulgaria and Romania, for
example, slowed the introduction of student loan programs, which are
important for raising tuition fees and diversifying the sources of tertiary
funding. Whatever measures countries in the region decide to adopt, it is
important that they continue to build learning assessment systems and
monitor and evaluate the impact of reforms on desired learning out-
comes. This approach offers the only way to determine whether reforms
have the intended effect, whether course corrections are needed, and
whether funding is achieving the ultimate goal of education systems in
the region: delivering the skills and competencies needed in the labor
market.
Notes
1. The PIRLS scale average is a metric established by setting the mean scores
of participating countries to 500 and the standard deviation to 100, thus
enabling comparisons over time, since all cycles are placed on this metric so
that scores are equivalent from cycle to cycle. In contrast, the international
average, obtained by averaging across the mean scores for each of the par-
ticipating countries, needs to be recomputed for each new cycle, based on
the set of participating countries and changes from cycle to cycle, depend-
ing on the set of countries taking part. For more information on PIRLS
2006, see its website at http://timss.bc.edu/pirls2006/index.html (accessed
January 2011).
2. The EBRD-World Bank Business Environment and Enterprise Performance
Surveys (BEEPS) provide successive snapshots of the self-reported constraints
to expansion by firms in the ECA region. The surveys were conducted in
virtually all ECA countries in 1999, 2002, 2005, and 2008. Survey samples
are constructed by random sampling from the national registry of firms (or
equivalent) and cover both industry and service sectors.
3. The Fourth European Work Conditions Survey conducted in 2005 covers the
EU-27 countries, Norway, Switzerland, Croatia, and Turkey (see Eurofound
2007).
4. The importance of effective teaching for learning is now increasingly recog-
nized. A good teacher can make a huge difference, with high-quality teachers
able to “get an entire year’s worth of additional learning out of their students
compared to those near the bottom” (Hanushek 2006, 3). That is, a good
teacher will get a gain of 1.5 grade-level equivalents, while a bad teacher will
get only the equivalent of 0.5 years for a single academic year. When consid-
ered over the entire school cycle, such differences accumulate rapidly and
make the impact of other variables small in comparison.
5. Although Turkey comes from a different tradition, education management
remains highly centralized. The impediments identified in this section—
specifically, too little information on student skills and competencies and too
little autonomy and accountability—apply with equal measure to Turkey.
References
Abromovsky, Laura, Erich Battistin, Emla Fitzsimons, Alissa Goodman, and Helen
Simpson. 2005. The Impact of the Employer Training Pilots on the Take-up of
Training Among Employers and Employees. London: Department for Education
and Skills, Government of the United Kingdom.
Acemoğlu, Daron. 1998. “Why Do New Technologies Complement Skills?
Directed Technical Change and Wage Inequality.” Quarterly Journal of
Economics 113: 1055–89.
Acosta, Pablo, and Leonardo Gasparini. 2007. “Capital Accumulation, Trade
Liberalization, and Rising Wage Inequality: The Case of Argentina.” Economic
Development and Cultural Change 55 (July): 793–812.
Adelman, Clifford. 2003. “Learning Accountability from Bologna: A Higher
Education Primer.” Institute for Higher Education Policy, Washington, DC.
AES (Adult Education Survey) (dataset). 2007. Eurostat, Luxembourg. http://
epp.eurostat.ec.europa.eu/portal/page/portal/microdata/adult_education_
survey.
Airola, Jim, and Chinhui Juhn. 2005. “Wage Inequality in Post-Reform Mexico.”
IZA Discussion Paper 1525, Institute for the Study of Labor, Bonn.
Allais, Stephanie. 2010. The Implementation and Impact of National Qualifications
Frameworks: Report of a Study in 16 Countries. Geneva: International Labour
Organization.
Alam, Asad, Mamta Murthi, Ruslan Yemtsov, Edmundo Murrugarra, Nora
Dudwick, Ellen Hamilton, and Erwin Tiongson. 2005. Growth, Poverty, and
225
226 References
Inequality: Eastern Europe and the Former Soviet Union. Washington, DC:
World Bank.
———. 2010. “When They Come to School, I Will Teach Them: When Equality
Leads to Inequality.” English translation of a Serbian-language PowerPoint
presentation delivered to the Serbian Academy of Sciences and Arts confer-
ence, “The Change of Roman Identity, Culture, and Language Conditioned by
Planned Socioeconomic Integration,” Belgrade, December 6–8.
Alam, Asad, Paloma Anós Casero, Faruk Khan, and Charles Udomsaph. 2008.
Unleashing Prosperity: Productivity Growth in Eastern Europe and the Former
Soviet Union. Washington, DC: World Bank.
Alvarez, Jesus V., Garcia Moreno, and Harry A. Patrinos. 2007. “Institutional
Effects as Determinants of Learning Outcomes: Exploring State Variations in
Mexico.” Well-Being and Social Policy 3 (1): 47–68.
Amaral, Alberto, Glen A. Jones, and Berit Karseth, eds. 2003. Governing Higher
Education: National Perspectives on Institutional Governance. New York:
Springer.
Autor, David H., Lawrence F. Katz, and Melissa S. Kearney. 2005. “Trends in U.S.
Wage Inequality: Re-Assessing the Revisionists.” NBER Working Paper 11627,
National Bureau of Economic Research, Cambridge, MA.
Autor, David H., Frank Levy, and Richard J. Murnane. 2003. “The Skill Content of
Recent Technological Change: An Empirical Exploration.” Quarterly Journal
of Economics 118 (4): 1279–334.
Barnow, Burt S. 2009. “Vouchers in U.S. Vocational Training Programs: An
Overview of What We Have Learned.” Zeitschrift für Arbeitsmarktforschung
42 (1): 71–84.
Barrera-Osorio, Felipe, Tazeen Fasih, Harry Anthony Patrinos. 2009. “Decentralized
Decision-Making in Schools: The Theory and Evidence of School-Based
Management.” Report 48740, World Bank, Washington, DC.
Bartelsman, Eric, John Haltivanger, and Stefano Scarpetta. 2004. “Microeconomic
Evidence of Creative Destruction in Industrial and Developing Countries.”
Policy Research Working Paper 3464, World Bank, Washington, DC.
Bartlett, William. 2007. “Economic Restructuring, Job Creation, and the Changing
Demand for Skills in the Western Balkans Area.” In Labour Markets in the
Western Balkans: Challenges for the Future, ed. Anastasia Fetsi, 19–50. Turin:
European Training Foundation.
Baucal, Aleksandar. 2006. “Development of Mathematics and Language Literacy
of Roma Students.” Psihologija 39 (2): 211–35.
———. 2010. “When They Come to School, I Will Teach Them: When Equality
Leads to Inequality.” English translation of a Serbian-language PowerPoint
presentation delivered to the Serbian Academy of Sciences and Arts confer-
References 227
ence, “The Change of Roman Identity, Culture, and Language Conditioned by

Planned Socioeconomic Integration,” Belgrade, December 6–8.
Baucal, Aleksandar, and Dragica Pavlović-Babić. 2009a. “Quality and Equity of
Education in Serbia: Educational Opportunities of the Vulnerable—PISA
Assessment, 2003 and 2006 Data.” Draft paper.
———. 2009b. “Kvalitet i pravednost obrazovanja u Srbiji: obrazovne šanse
siromašnih.” Ministarstvo prosvete Republike Srbije i Institut za psihologiju,
Beograd.
BEEPS (Business Environment and Enterprise Performance Survey) (dataset). 2005,
2008. EBRD (European Bank for Reconstruction and Development), London,
and World Bank, Washington, DC. http://go.worldbank.org/I4ZRB5O300.
Behn, Robert D. 2001. Rethinking Democratic Accountability. Washington, DC:
Brookings Institution.
Berman, Eli, John Bound, and Stephen Machin. 1998. “Implications of Skill-
Biased Technological Change: International Evidence.” Quarterly Journal of
Economics 113 (4): 1245–79.
Berman, Eli, and Stephen Machin. 2000. “Skill-Biased Technology Transfer
Around the World.” Oxford Review of Economic Policy 16 (3): 12–22.
Betcherman, Gordon, Karina Olivas, and Amit Dar. 2004. “Impact of Active Labor
Market Programs: New Evidence from Evaluations with Particular Attention
to Developing and Transition Countries.” Social Protection Discussion Paper
0402, World Bank, Washington, DC.
Bologna Working Group on Qualifications Frameworks. 2005. “A Framework for
Qualifications of the European Higher Education Areas.” Danish Ministry of
Science Technology and Innovation, Copenhagen.
Bound, John, and George Johnson. 1995. “What are the Causes of Rising Wage
Inequality in the United States?” Economic Policy Review 1 (1): 9–17.
Bourguignon, François, and F. Halsey Rogers. 2008. “Global Returns to Higher
Education: Trends, Drivers, and Policy Responses.” Keynote address at the
Annual World Bank Conference on Development Economics, Cape Town,
South Africa, June 9–11.
Breshnan, Timothy F., Erik Brynjolfsson, and Lorin M. Hitt. 2002. “Information
Technology, Workplace Organization, and the Demand for Skilled Labor:
Firm Level Evidence.” Quarterly Journal of Economics 117 (1): 339–76.
Bulgaria, Government of. 2007. Raw dataset from Bulgaria 2007 Labor Force
Survey. National Statistical Institute, Sofia, Bulgaria.
Brunner, Jose Joaquin, and Anthony Tillett. 2007. “Higher Education in Central
Asia: The Challenge of Modernization.” World Bank, Washington, DC.
Campos, Nauro, and Aurelijus Dabušinskas. 2008. “So Many Rocket Scientists,
So Few Marketing Clerks: Estimating the Effects of Economic Reform on
228 References
Occupational Mobility in Estonia.” IZA Discussion Paper 3886, Institute for

the Study of Labor, Bonn.
Campos, Nauro, and Dean Jolliffe. 2007. “Earnings, Schooling, and Economic
Reform: Econometric Evidence from Hungary (1986–2004).” World Bank
Economic Review 21 (3): 509–26.
Campos, Nauro, and Dana Žlábková. 2001. “The Wrong Mix: A First Look at
Occupational Mobility during the Hungarian Transition.” Discussion Paper
56, Center for Economic Research and Graduate Education-Economics
Institute (CERGE-EI DP), Prague.
Card, David, Jochen Kluve, and Andrea Weber. 2009. “Active Labor Market Policy
Evaluations: A Meta Analysis.” IZA Discussion Paper 4002, Institute for the
Study of Labor, Bonn.
Carey, David. 2007. “Improving Education Outcomes in the Slovak Republic.”
OECD Economics Department Working Paper 578, Organisation for
Economic Co-operation and Development, Paris.
Carey, Kevin. 2010. “That Old College Lie: Are Our Colleges Teaching Students
Well? No. But Here’s How to Make Them.” Democracy 15 (Winter): 8–20.
Carnoy, M., and S. Loeb. 2002. “Does External Accountability Affect Student
Outcomes? A Cross-state Analysis.” Educational Evaluation and Policy
Analysis 24 (4): 305–31.
CEDEFOP (European Centre for the Development of Vocational Training).
2009. “Using Tax Incentives to Promote Education and Training.” Cedefop,
Thessaloniki.
Chawla, Mukesh, Gordon Betcherman, and Arup Banerji. 2007. From Red to
Gray: The “Third Transition” of Aging Populations in Eastern Europe and the
Former Soviet Union. Washington, DC: World Bank.
CIE (Center for Innovations in Education). 2009. “Math and Reading Skills of
Azeri Students: Possible Reasons for Relatively High Performance in Math
and Poor Performance in Reading in PISA 2006.” Draft report prepared
for the World Bank, CIE, Baku. Text available on the Curriculum Portal,
Ministry of Education of the Republic of Azerbaijan, Baju, http://www
.kurikulum.az/files/Analysis%20of%20Azerbaijan%27s%20performance%
20in%20PISA%202006%20%28Eng%29.pdf.
Coalition for Clean Universities. 2009: “The University Integrity Contest—An
Integrity System for the Romanian Higher Education.” Online publication of
the Romanian Academic Society (SAR), Bucharest. http://www.sar.org.ro/
files/366_CUC%20report.pdf.
Commander, Simon, and János Köllö. 2004. “The Changing Demand for Skills:
Evidence from the Transition.” IZA Discussion Paper 1073, Institute for the
Study of Labor, Bonn.
References 229
Contreras, D., L. Flores, and F. Lobato. 2003. “Monetary Incentives for Teachers
and School Performance: The Evidence for Chile.” Department of Economics,
University of Chile, Santiago.
Crostat (Republic of Croatia Central Bureau of Statistics). 2007. Raw dataset
from Croatia 2007 Labor Force Survey, Crostat, Zagreb, Croatia.
CVTS (Continuing Vocational Training Survey) (dataset). 1995, 2005. Eurostat,
Luxembourg. http://epp.eurostat.ec.europa.eu/portal/page/portal/education/
introduction.
Czech Republic, Government of. 2007. “The Strategy of Lifelong Learning in the
Czech Republic.” Government Decree 761/2007, Prague.
Dahlman, Carl. 2009. “Different Innovation Strategies, Different Results: Brazil,
Russia, India, China, and Korea (the BRICKs).” In Innovation and Growth:
Chasing a Moving Frontier, ed. Vandana Chandra, Deniz Eröcal, Pier C.
Padoan, and Carlos A. P. Braga, 131–68. Paris: Organisation for Economic
Co-operation and Development and World Bank.
Dar, Amit, Sudharshan Canagarajah, and Paud Murphy. 2003. “Training Levies:
Rationale and Evidence from Evaluations.” World Bank, Washington, DC.
Dearden, Lorraine, Leslie McGranahan, and Barbara Sianesi. 2004. An In-Depth
Analysis of the Returns to National Vocational Qualifications Obtained at Level
2. London: Centre for the Economics of Education.
Dearden, Lorraine, Howard Reed, and John van Reenen. 2000. “Who Gains
When Workers Train? Training and Corporate Productivity in a Panel of
British Industries.” CEPR Discussion Paper 2486, Centre for Economic Policy
Research, London.
Denhardt, Janet V., and Robert B. Denhardt. 2003. The New Public Service:
Serving, Not Steering. New York: M. E. Sharpe.
Duke, Chris, Abrar Hasan, Paul Cappon, Werner Meissner, Hilary Metcalf, and
Don Thornhill. 2008. OECD Reviews of Tertiary Education: Croatia. Paris:
Organisation for Economic Co-operation and Development.
EBRD (European Bank for Reconstruction and Development)–World Bank.
2006. “Life in Transition Survey 2006.” EBRD, London, and World Bank,
Washington, DC. Data available on the EBRD website, http://www.ebrd.com/
pages/research/analysis/publications/transition/data.shtml.
EC (European Commission). 2008. “New Skills for New Jobs. Anticipating and
Matching Labour Market and Skill Needs.” Commission Staff Working
Document accompanying the Communication from the Commission to the
European Parliament, the Council, the European Economic and Social
Committee, and the Committee of the Regions. EC, Brussels.
Edstats (database). World Bank, Washington, DC. http://go.worldbank.org/
ITABCOGIV1.
230 References
Ercan, Hakan. 2008. “The Evolution of Demand for Skills in Turkey.” Background
paper for Skills—Not Just Diplomas: Managing for Results in Education
Systems in Eastern Europe and Central Asia, ed. Lars Sondergaard and Mamta
Murthi (Washington, DC: World Bank, forthcoming). Unpublished; available
upon request.
Estermann, Thomas, and Terhi Nokkala. 2009. University Autonomy in Europe I.
Exploratory Study. Brussels: European University Association.
ETF (European Training Foundation). 2007. “Kyrgyzstan: ETF Country Plan
2007.” ETF, Turin.
EU (European Union). 2008. “Recommendation of the European Parliament
and of the Council of 23 April 2008 on the Establishment of the European
Qualifications Framework for Lifelong Learning.” Official Journal of the European
Union C (111): 1–7. http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=
oj:c:2008:111:0001:0007:en:pdf.
Eurofound (European Foundation for the Improvement of Living and Working
Conditions). 2007. “Fourth European Working Conditions Survey” (EWCS
2005). Eurofound, Dublin, Ireland. The EWCS 2005 dataset is accessible
via the UK Data Archive (UKDA) of the University of Essex at www.esds
.ac.uk.
European Integration Consortium. 2009. “Labour Mobility within the EU in the
Context of Enlargement and the Functioning of the Transitional Arrangements.”
Employment, Social Affairs, and Equal Opportunities Directorate General,
European Commission, Nuremberg.
Eurostat SD (Statistics Database). Education and Training. Eurostat, Luxembourg.
http://epp.eurostat.ec.europa.eu/portal/page/portal/education/data/database.
———. Labour Market. Eurostat, Luxembourg. http://epp.eurostat.ec.europa.eu/
portal/page/portal/labour_market/introduction.
Fielden, J. 2008. “Global Trends in University Governance.” Education Working
Paper Series 9. Education Unit, Human Development Network, World Bank,
Washington, DC.
Flabbi, Luca, Stefano Paternostro, and Erwin R. Tiongson. 2007. “Returns to
Education in the Economic Transition: A Systematic Assessment Using
Comparable Data.” Policy Research Working Paper 4225, World Bank,
Washington, DC.
Fleisher, Belton M., Klara Sabirianova, and Xiaojun Wang. 2005. “Returns to Skills
and the Speed of Reforms: Evidence from Central and Eastern Europe, China,
and Russia.” Journal of Comparative Economics 33 (2): 351–70.
Glazerman, Steven, Eric Isenberg, Martha Bleeker, Amy Johnson, Julieta
Lugo-Gil, Mary Grider, and Sarah Dolfin. 2009. “Impacts of Comprehensive
Teacher Induction: Results from the First Year of a Randomized Controlled
References 231
Study.” NCEE 2009-4072, National Center for Education and Evaluation,

U.S. Department of Education, Washington, DC.
Godfrey, Martin. 2003. “Youth Employment Policy in Developing and Transition
Countries—Prevention as well as Cure.” Social Protection Discussion Paper
Series 0320, World Bank, Washington, DC.
Gorodnichenko, Yuriy, and Klara Sabirianova Peter. 2005. “Returns to Schooling
in Russia and Ukraine: A Semi Parametric Approach to Cross-Country
Comparative Analysis.” Journal of Comparative Economics 33 (2): 324–50.
Gunny, Madeleine, and Evelyn Viertel. 2007. Designing Adult Learning Strategies:
the Case of South Eastern Europe. Turin: European Training Foundation.
Hanushek, Eric A. 2006. “The Single Salary Schedule and Other Issues of Teacher
Pay.” Paper prepared for the University of Arkansas Department of Education
Reform Technical Board of Advisors Conference Agenda, Fayetteville,
October 19–21.
Hanushek, Eric A., and M.E. Raymond. 2005. “Does School Accountability
Lead to Improved Student Performance?” Journal of Policy Analysis and
Management 24 (2): 297–327.
Hanushek, Eric A., and Ludger Wößmann. 2008. “The Role of Cognitive Skills in
Economic Development.” Journal of Economic Literature 46 (3): 607–68.
Harmon, Robert, and Keith MacAllum. 2003. “Documented Characteristics of
Labor Market-Responsive Community Colleges and a Review of Supporting
Literature.” U.S. Department of Education, Washington, DC.
Hazan, Mihails. 2009. “Latvian Teachers: Underpaid and Over-Regulated?”
Background study for the World Bank project, “Latvia Development Policy
Lending: Social Sectors and Public Administration Reform,” World Bank,
Washington, DC.
Herbst, Mikołaj, Jan Herczyński, and Anthony Levitas. 2009. Finansowanie
Oświaty w Polsce—Diagnoza, Dylematy, Możliwości (Financing of Education
in Poland—Diagnosis, Dilemmas, Opportunities).” Warsaw: Wydanie
Pierwsze.
Heckman, James J. 1999. “Policies to Foster Human Capital.” NBER Working
Paper 7288, National Bureau of Economic Research, Cambridge, MA.
Heyneman, Stephen P., Kathryn H. Anderson, and Nazym Nuraliyeva. 2008. “The
Cost of Corruption in Higher Education.” Comparative Education Review
52 (1): 1–25.
Holdsworth, Nick. 2009. “Czech Republic: Degree Audit Follows Corruption
Claims.” University World News, November 8.
Huitfeldt, Henrik, Jens Johansen, and Irena Kogan. 2008. “An Analysis of Labour
Market Entry in Serbia and Ukraine.” In Transition from Education to Work in
EU Neighboring Countries. Turin: European Training Foundation.
232 References
Hungary, Ministry of Education and Culture. 2006. “Recognition of Non-

Formal and Informal Learning; OECD RNFIL Project: Country Background
Report—Hungary.” Ministry of Education and Culture, Budapest.
IEA (International Association for the Evaluation of Educational Achievements).
1995, 1999, 2003, and 2007. “Mathematics Teacher Background Data
Almanac.” In “Trends in International Mathematics and Science Study
(TIMSS).” TIMSS International Study Center, Boston College, Chestnut
Hill, MA.
———. 1995, 1999, 2003, and 2007. “Science Teacher Background Data
Almanac.” In “Trends in Mathematics and Science Study (TIMSS).” TIMSS
International Study Center, Boston College, Chestnut Hill, MA.
IES (International Center for Data Statistics). 2001, 2006. “PIRLS (Progress
on International Reading Literacy Study) Results.” IES, U.S. Department
of Education Institute of Education Sciences, Washington, DC. http://nces
.ed.gov/surveys/pirls/.
IFF Research. 2008. “Understanding the Impact of the Adult Learning Grant.”
Learning and Skills Council, London.
ILO (International Labour Organization). Bureau of Statistics. 1988. International
Standard Classification of Occupations (ISCO-88). 3rd rev. ILO: Geneva. http://
www.ilo.org/public/english/bureau/stat/isco/isco88/index.htm.
Ireland, National Qualifications Authority of. 2003. “The Role of National
Qualification Systems in Promoting Lifelong Learning. Background Report
for Ireland.” OECD, Paris.
Ivaschenko, Oleksiy. 2008. “The Survey of Skills, Labor Demand and Job
Vacancies in Kazakhstan. Main Findings.” PowerPoint presentation, European
and Central Asia Region, World Bank, Washington, DC, March 11.
Jakubowski, Maciej, Harry Anthony Patrinos, Emilio Ernesto Porta, and Jerzy
Wiśniewski. 2010. “The Impact of the 1999 Education Reform in Poland.”
Policy Research Working Paper 5263, World Bank, Washington, DC.
Jongbloed, Ben. 2008. “Strengthening Consumer Choice in Higher Education.”
In Cost-Sharing and Accessibility in Higher Education: A Fairer Deal? ed. Pedro
N. Teixeira, D. Bruce Johnstone, Maria J. Rosa, and J. J. Vossensteijn, 19–50.
New York: Springer.
Juhn, Chinhui, Kevin M. Murphy, and Brooks Pierce. 1993. “Wage Inequality and
the Rise in Returns to Skill.” Journal of Political Economy 101 (3): 410–42.
Kang, Seoghoon, and Dong-Pyo Hong. 2002. “Technological Change and Demand
for Skills in Developing Countries: An Empirical Investigation of the Republic
of Korea’s Case.” The Developing Economies 40 (2): 188–207.
Kapelyushnikov, Rostislav. 2008. “Russia’s Human Capital: An Assessment.”
Background Paper for Skills—Not Just Diplomas: Managing for Results in
Education Systems in Eastern Europe and Central Asia, ed. Lars Sondergaard and
References 233
Mamta Murthi (Washington, DC: World Bank, forthcoming). Unpublished;

available upon request.
Kehm, Barbara M., and Bjørn Stensaker, eds. 2009. University Rankings, Diversity,
and the New Landscape of Higher Education. Global Perspectives on Higher
Education, Vol. 18. Rotterdam: Sense Publishers.
Kheyfets, Igor, Massimo Mastruzzi, Dino Merotto, and Lars Sondergaard. 2011. “A
New Data Tool to BOOST Public Spending Efficiency.” Europe and Central
Asia Knowledge Brief Vol. 43 (September 2011), World Bank, Washington,
DC.
Korka, Mihai. 2008. “Implementing the Modern Approach of Quality Assurance in
Romanian Higher Education.” Paper presented at the “Income Convergence”
Conference, Bucharest, April 8.
Kouznetsova, Tatiana. 2009. “Strengths and Weaknesses of Reading Activity of
Primary School Graduates in Russia according to PIRLS: 2006 Results.”
Education Issues (1) [State University Higher School of Economics,
Moscow].
Krstić, Gorana, and Peter Sanfey. Forthcoming. “Earnings Inequality and the
Informal Economy: Evidence from Serbia.” Economics of Transition.
Kertesi, Gabor, and Janos Kollo. 1999. “Economic Transformation and the Return
to Human Capital. The Case of Hungary, 1986-1996.” Budapest Working
Paper on the Labour Market 9907, Institute of Economics, Hungarian
Academy of Sciences, Budapest.
———. 2001. “Economic Transformation and the Revaluation of Human Capital
—Hungary 1986–1999.” Budapest Working Paper on the Labour Market
0104, Institute of Sciences, Hungarian Academy of Sciences, Budapest.
Leinbach, D. Timothy, and Davis Jenkins. 2008. “Using Longitudinal Data to
Increase Community College Student Success: A Guide to Measuring
Milestone and Momentum Point Attainment.” CCRC (Community College
Research Center) Research Tools 2 (January).
Leitch Review of Skills. 2006. Prosperity for All in the Global Economy: World Class
Skills. London: HM Treasury.
Levy, Frank, and Richard J. Murnane. 2004. The New Division of Labor: How
Computers are Creating the Next Job Market. Princeton: Princeton University
Press.
LFS (Labour Force Survey) (database). Eurostat, Luxembourg. http://epp.euro
stat.ec.europa.eu/portal/page/portal/labour_market/introduction.
Linden, Tony, Nina Arnhold, and Kirill Vasiliev. 2008. “From Fragmentation
to Cooperation: Tertiary Education, Research, and Development in South
Eastern Europe.” Education Working Paper 13, World Bank, Washington, DC.
Martin, Rob, Frank Villeneuve-Smith, Liz Marshall, and Ewan McKenzie. 2008.
Employability Skills Explored. London: Learning and Skills Network.
234 References
Machin, Stephen, and John Van Reenen. 1998. “Technology and Changes in
Skill Structure: Evidence from Seven OECD Countries.” Quarterly Journal of
Economics 113 (4): 1215–44.
Mansoor, Ali, and Bryce Quillin, eds. 2007. Migration and Remittances: Eastern
Europe and the Former Soviet Union. Washington, DC: World Bank.
McKinsey & Company. 2007. “How the World’s Best-Performing School Systems
Come Out on Top.” McKinsey & Company, Social Sector Office, New York.
http://www.mckinsey.com/clientservice/Social_Sector/our_practices/
Education/Knowledge_Highlights.aspx.
———. 2008: “An Introduction to the McKinsey Report and What It Means
for Education in Australia.” Presentation, McKinsey & Company, New York.
Cited with permission.
———. 2010. “Closing the Talent Gap: Attracting and Retaining Top-Third
Graduates to Careers in Teaching—An International and Market Research-
Based Perspective” Byron Auguste, Paul Kihn, and Matt Miller, McKinsey
& Company, Social Sector Office, New York. http://www.mckinsey.com/
clientservice/Social_Sector/our_practices/Education/Knowledge_Highlights
.aspx.
Mitra, Pradeep, Marcelo Selowsky, and Juan Zalduendo. 2010. Turmoil at Twenty:
Recession, Recovery, and Reform in Central and Eastern Europe and the Former
Soviet Union. Washington, DC: World Bank.
Mungiu-Pippidi, Alina. 2009. “Assessing Corruption in the Romanian Higher
Education System: The Coalition for Clean Universities.” Photocopy, Hertie
School of Governance, Berlin.
Munich, Daniel, Jan Svejnar, and Katherine Terrell. 1999. “Returns to Human
Capital under the Communist Wage Grid During the Transition to a Market
Economy.” Review of Economics and Statistics 87 (1): 100–127.
Nastasescu, Vasile. 2006. “Romania: Higher Education System and Its Quality
Assurance.” Paper presented at The Third Europe and Central Asia (ECA)
Education Conference, “Quality and Relevance of Education,” St. Petersburg,
October 4–6.
Newell, Andrew, and Barry Reilly. 1997. “Rates of Returns to Educational
Qualifications in the Transitional Economies.” Discussion Paper in Economics
03/97, University of Sussex, Brighton.
Newell, Andrew, and Mieczyslaw W. Socha. 2007. “The Polish Wage Inequality
Explosion.” IZA Discussion Paper 2644, Institute for the Study of Labor,
Bonn.
OECD (Organisation for Economic Co-operation and Development). 1994. The
OECD Jobs Study: Facts, Analysis, Strategies. Paris: OECD.
———. 1999a. Measuring Student Knowledge and Skills: A New Framework for
Assessment. Paris: OECD.
References 235
———. 1999b. OECD Employment Outlook: Giving Youth a Better Start. Paris:
OECD.
———. 2004a. “Improving Skills for More and Better Jobs: Does Training Make a
Difference?” In OECD Employment Outlook 2004, 183–224. Paris: OECD.
———. 2004b. “Raising the Quality of Educational Performance at School.” Policy
Brief, OCED, Paris. http://www.oecd.org/publications/Pol_brief.
———. 2005a. Attracting, Developing, and Retaining Effective Teachers—Final
Report: Teachers Matter. Paris: OECD.
———. 2005b. “The Definition and Selection of Key Competencies—Executive
Summary.” OECD, Paris.
———. 2005c. Modernizing Government: The Way Forward. Paris: OECD.
———. 2005d. Promoting Adult Learning. Paris: OECD.
———. 2005e. “Thematic Review of Adult Learning: Poland Country Note.”
OECD, Paris.
———. 2006. “OECD Economic Surveys: Netherlands.” OECD, Paris.
———. 2007a. Analysis. Volume 1 of PISA 2006: Science Competencies for
Tomorrow’s World. Paris: OECD.
———. 2007b. Education at a Glance: OECD Indicators. Paris: OECD.
———. 2007c. Performance-Based Budgeting in OECD Countries. Paris: OECD.
———. 2007d. “PISA 2006: Science Competencies for Tomorrow’s World;
Executive Summary.” OECD, Paris.
———. 2008. Education at a Glance 2008: OECD Indicators. Paris: OECD.
OECD Education At a Glance Database. OECD, Paris. http://www.oecd.org/doc
ument/30/0,3343,en_2649_39263238_39251550_1_1_1_1,00.html.
OECD and World Bank. 2007. Reviews of National Policies for Education: Higher
Education in Kazakhstan. Paris: OECD.
Ok, Wooseok, and Peter Tergeist. 2003. “Improving Workers’ Skills: Analytical
Evidence and the Role of the Social Partners.” OECD Social, Employment,
and Migration Working Papers 10, Organisation for Economic Co-operation
and Development, Paris.
Orazem, Peter, and Milan Vodopivec. 1994. “Winners and Losers in Transition:
Returns to Education, Experience, and Gender in Slovenia.” Policy Research
Working Paper 1342, World Bank, Washington, DC.
Osborne, David, and Ted Gaebler. 1992. Reinventing Government: How the
Entrepreneurial Spirit is Transforming the Public Sector. Reading, MA: Addison-
Wesley Publishing Company, Inc.
OSI (Open Society Institute). 2007. Equal Access to Quality Education for Roma.
Vol. 1. Monitoring Report. Budapest: OSI/EU Monitoring and Advocacy
Program.
236 References
Pachuashvili, Maria. 2007. “Changing Patterns of Private-Public Growth and

Decline: The Case of Georgian Higher Education.” PROPHE Working Paper
10, Program for Research on Private Higher Education, Albany, NY.
Patrinos, Harry Anthony. 2008. “Labor Markets and Education and Training in
Ukraine.” Photocopy, World Bank, Washington DC.
Patrinos, Harry Anthony, Tazeen Fasih, Juliana Guaqueta, Emilio Porta, Kevin
Macdonald, and Plamen Danchev. 2010. “A Review of the Bulgaria School
Autonomy Reforms.” Report 54890-BG, World Bank, Washington, DC.
Peter, Klara Sabirianova. 2003. “Skill-Biased Transition: The Role of Markets,
Institutions, and Technological Change.” IZA Discussion Paper 893, Institute
for the Study of Labor, Bonn.
PISA Database. 2000, 2003, 2006, 2009. Organisation for Economic Co-operation
and Development, Paris. http://www.oecd.org/edu/pisa/2009.
Poland, Government of. 2007. Raw dataset from Poland 2007 Labor Force Survey,
Central Statistical Office, Warsaw.
Poland, National Bank of. 2008. “Yearly Survey of Enterprises, 2008.” National
Bank of Poland, Warsaw.
Pricopie, Remus, Valeriu Frunzary, Nicoleta Corbu, and Ivan Loredana. 2010.
“Arguments for a New Policy Dialog on Access and Equity.” Romanian Journal
of Communication and Public Relations 12 (2): 9–25.
Psacharopoulos, George, and Harry A. Patrinos. 2004. “Returns to Investment in
Education: A Further Update.” Education Economics 12 (2): 111–34.
Rauhvargers, Andrejs, Cynthia Deane, and Wilfried Pauwels. 2009. “Bologna Process
Stocktaking Report 2009.” Report from working groups appointed by the
Bologna Follow-up Group to the Ministerial Conference in Leuven/Louvain-
la-Neuve, Lifelong Learning Programme, European Commission, Brussels.
Romania, Government of. National Institute of Statistics. 2008. Romanian
Statistical Yearbook 2008. Bucharest: National Institute of Statistics.
———, Ministry of Education, Research and Innovation. 2009. “Report on
the State of the National Education System 2009.” Ministry of Education,
Research and Innovation, Bucharest.
Russian Federation, Government of. 2007. Raw dataset from 2007 Russian
Federation Labor Force Survey, Federal State Statistics Service, Moscow.
Rutkowski, Jan. 1996a. “Changes in the Wage Structure during Economic
Transition in Central and Eastern Europe.” World Bank Technical Paper 340,
World Bank, Washington, DC.
———. 1996b. “High Skills Pay-off: The Changing Wage Structure during
Economic Transition in Poland.” Economics of Transition 4 (1): 89–112.
———. 2001. “Earnings Inequality in Transition Economies of Central Europe:
Trends and Patterns During the 1990s.” Social Protection Discussion Paper
0117, World Bank, Washington, DC.
References 237
———. 2007. “From the Shortage of Jobs to the Shortage of Skilled Workers.
Labor Markets in the EU New Member States.” IZA Discussion Paper 3202,
Institute for the Study of Labor, Bonn.
———. 2008a. “Does the Labor Market Support Croatia’s EU Convergence?”
Photocopy, World Bank, Washington, DC. Unpublished background paper
for World Bank, 2009, “Croatia’s EU Convergence Report: Reaching and
Sustaining Higher Rates of Economic Growth.” Report 48879-HR, 2 vols.,
Europe and Central Asia Region, World Bank, Washington, DC.
———. 2008b. “Labor Market in Georgia: Lack of Jobs or Structural Mismatches?”
Photocopy, World Bank, Washington, DC. Unpublished background paper for
Georgia Country Economic Memorandum, World Bank, Washington, DC.
———. 2010. “Demand for Skills in FYR Macedonia.” Photocopy. Report sup-
ported by the Multidonor Trust Fund, “Labor Markets, Job Creation, and
Economic Growth,” through funding of the German, Norwegian, Austrian,
and Korean governments. World Bank, Washington, DC.
Rutkowski, Jan, and Stefano Scarpetta. 2005. Enhancing Job Opportunities: Eastern
Europe and the Former Soviet Union. With Arup Banerji, Philip O’Keefe,
Gaëlle Pierre, and Milan Vodopivec. Washington, DC: World Bank.
Salmi, Jamil. 2009. “The Growing Accountability Agenda in Tertiary Education:
Progress or Mixed Blessing?” World Bank Education Working Paper Series 16.
Document 47760, Education Unit, Human Development Network, World
Bank, Washington, DC.
———. 2011. “Review of Student Learning Assessment Experiences in Higher
Education.” Unpublished paper, World Bank, Washington, DC.
Santiago, Paulo, and Karine Tremblay. 2008. “Setting the Right Course, Steering
Higher Education.” With Marie-Claire Duguay and Thomas Weko. In Special
Features: Governance, Funding, Quality, Volume 1 of Tertiary Education for the
Knowledge Society, P. Santiago, K. Tremblay, E. Basri, and E. Arnal, 67–162.
Paris: OECD.
Scarpetta, Stefano, and Milan Vodopivec. 2005. “Restructuring, Productivity,
and Job Creation.” In Enhancing Job Opportunities: Eastern Europe and the
Former Soviet Union, ed. Jan J. Rutkowski and Stefano Scarpetta, 125–54.
Washington, DC: World Bank.
Silova, Iveta. 2009. “The Crisis of the Post-Soviet Teaching Profession in the
Caucasus and Central Asia.” Research in Comparative and International
Education 4 (4): 367–84.
Slantcheva, Snejana, and Daniel Levy, eds. 2007. The Search for Legitimacy:
Private Higher Education in Central and Eastern Europe. Basingstoke: Palgrave-
MacMillan.
Steiner-Khamsi, Gita, and Christine Harris-Van Keuren. 2008. “The Pendulum
of Decentralization and Recentralization Reforms: Its Impact on Teacher
238 References
Salaries in the Caucasus, Central Asia, and Mongolia.” With Iveta Silova
and Ketevan Chachkhiani. Background paper for Education for All Global
Monitoring Report 2009: Overcoming Inequality: Why Governance Matters.
Paris: UNESCO; London: Oxford University Press.
Steiner-Khamsi, Gita, and Iveta Silova. 2007. How NGOs React: Globalization
and Education Reform in the Caucasus, Central Asia and Mongolia. West
Hartfield, CT: Kumarian Press.
Stensaker, Bjørn, Jürgen Enders, and Harry de Boer. 2006. “Comparative
Analysis and Executive Summary.” Part One of “The Extent and Impact
of Higher Education Governance Reform Across Europe.” Final report
to the Directorate-General for Education and Culture of the European
Commission, Center for Higher Education and Policy Studies, University of
Twente, Enschede, The Netherlands.
Tan, Hong. 2001. “Do Training Levies Work? Malaysia’s HRDF and Its Effects on
Training, and Firm-Level Productivity.” Working Paper, World Bank Institute,
Washington, DC.
Tan, Hong, and Gladys Lopez Acevedo. 2005. “Evaluating Training Programs
for Small and Medium Enterprises, Lessons from Mexico.” Policy Research
Working Paper 3760, World Bank, Washington DC.
Tan, Hong, Yevgeniya Savchenko, Vladimir Gimpelson, Rostislav Kapelyushnikov,
and Anna Lukyanova. 2007. “Skills Shortages and Training in Russian Enter-
prises.” Policy Research Working Paper 4222, World Bank, Washington, DC.
Tumeneva, Yulia. 2006. “Analysis of Russian Education Features as Correlates
with Student Achievements in PISA.” In Russian Schools: From PISA 2000 to
PISA 2003, ed. Anatolii Kasprjak, 110–76. Moscow: Logos.
UIS (UNESCO Institute for Statistics). Data Centre (various databases).
UIS. Montreal. http://stats.uis.unesco.org/unesco/TableViewer/document
.aspx?ReportId=143&IF_Language=eng.
UNECE (United Nations Commission for Europe) SD (Statistical Database),
Labour Force and Wages. UNECE, Geneva, Switzerland. http://w3.unece
.org/pxweb/database/STAT/20-ME/3-MELF/?lang=1.
UNICEF (United Nations Children’s Fund). 2007. Education for Some More than
Others? Geneva: UNICEF.
United Kingdom, DFES (Department for Education and Skills). 2003. “The
Role of National Qualification Systems in Promoting Lifelong Learning.
Background Report for the United Kingdom.” Organisation for Economic
Co-operation and Development, Paris.
United Nations. 2006. “World Population Prospects—the 2006 Revision.”
Department of Economic and Social Affairs, Population Division, United
Nations, New York.
References 239
Usher, Alex. 2009. “Ten Years Back and Ten Years Forward: Developments and
Trends in Higher Education in Europe Region.” Paper presented at the
UNESCO Forum on “Higher Education in the Europe Region: Access, Values,
Quality and Competitiveness,” Bucharest, May 21–24.
Usher, Alex, and Jon Medow. 2009. “A Global Survey of University Rankings
and League Tables.” In University Rankings, Diversity, and the New Landscape
of Higher Education, ed. Barbara M. Kehm and Bjørn Stensaker, 3–18.
Vol. 18 of Global Perspectives on Higher Education. Rotterdam: Sense
Publishers. The university rankings of the Center for Higher Education
Development (CHE) in Gütersloh, Germany, can be accessed at http://www
.che-ranking.de/cms/?getObject=615&getLang=en.
Vegas, Emiliana, and Jenny Petrow. 2008. Raising Student Learning in Latin
America: the Challenge for the 21st Century. Washington, DC: World Bank.
Vincze, Maria, and Janos Kollo. 1999. “Economic Transformation and the Return
to Human Capital. The Case of Hungary, 1986–1996.” Budapest Working
Papers on the Labour Market 9907, Institute of Economics, Hungarian
Academy of Sciences, Budapest.
World Bank. 2003a. Raw dataset from 2003 Tajikistan Living Standards
Measurement Survey, World Bank, Washington, DC.
———.2003b. “Project Appraisal Document on a Proposed Loan in the Amount
of $60 Million to Romania for a Rural Education Project.” Report 25101-RO,
World Bank, Washington, DC.
———. 2003c. The World Development Report 2004: Making Services Work.
Washington, DC: World Bank.
———. 2005a. Expanding Opportunities and Building Competencies for Young
People. A New Agenda for Secondary Education. Washington, DC: World
Bank.
———. 2005b. Raw dataset from 2005 Uzbekistan Regional Panel Survey, World
———. 2007a. “Azerbaijan Efficiency Issues in General Education: Results from
a Public Expenditure Tracking and Service Delivery Survey (PETSDS) of
General Education in Azerbaijan.” Draft report, World Bank, Washington, DC.
———. 2007b. “Accelerating Bulgaria’s Convergence: The Challenge of Raising
Productivity.” Report 38570, Poverty Reduction and Economic Management
Unit, World Bank, Washington, DC.
———. 2007c. “The Labor Market and Education and Training in the Western
Balkans.” Draft policy note, World Bank, Washington, DC. http://siteresources
.worldbank.org/EXTECAREGTOPEDUCATION Resources/444607-11926
36551820/Labor_Market_and_Education_Note_Western_Balkans
.June_28a.pdf.
240 References
———. 2007d. “Turkey—Higher Education Policy Study.” Report 9674, Human

Development Sector Unit, World Bank, Washington, DC.
———. 2008a. “Czech Republic: Improving Employment Chances for the Roma.”
Report 46120, Human Development Sector, World Bank, Washington, DC.
———. 2008b. “Introducing a Student Loan Scheme in Romania. A Discussion
Paper.” Report 46206, World Bank, Washington, DC.
———. 2008c. “Investing in Turkey’s Next Generation: The School-to-Work
Transition and Turkey’s Development.” Report 44048, Human Development
Sector Unit, World Bank, Washington, DC.
———. 2008d. “Ukraine—Improving Intergovernmental Fiscal Relations and
Public Health and Education Expenditures Policy: Selected Issues.” Report
42450, Poverty Reduction and Economic Management Unit, Europe and
Central Asian Region, World Bank, Washington, DC.
———. 2009a. “Croatia’s EU Convergence Report: Reaching and Sustaining
Higher Rates of Economic Growth.” 2 vols. World Bank, Washington, DC.
———. 2009b. “Slovak Public Expenditure Review.” World Bank, Washington,
DC.
———. 2009c. “Ukraine Labor Demand Study.” Human Development Sector
Unit, Europe and Central Asia Region, World Bank, Washington, DC.
———. 2010a. “Azerbaijan: Living Conditions Assessment Report.” Report
52801, Human Development Sector Region, World Bank, Washington, DC.
———. 2010b. “Europe 2020—Poland; Fueling Growth and Competitiveness in
Poland—Through Employment, Skills, and Innovation.” Draft report, World
———. 2011a. “Improving the Quality and Equity of Basic Education in Turkey—
Challenges and Options.” Policy Note 54131-TR, World Bank, Washington,
DC.
———. 2011b. “Strengthening Skills and Employability in Peru.” Report No.
61699-PE, World Bank, Washington, DC.
World Development Indicators (database). World Bank, Washington, DC. http://
data.worldbank.org/data-catalog/world-development-indicators.
Yemtsov, Ruslan G., Stefania R. Cnobloch, and Cem Mete. 2006. “Evolution of
the Predictors of Earnings during Transition.” Unpublished manuscript, World
Zeigele, Frank. 2009. “How to Ensure Equity, Quality and Efficiency: Performance-
Based Budgeting.” Presentation delivered at the Fifth World Bank ECA
Education Conference, Budva, Montenegro, October 27–29.
ECO-AUDIT
Environmental Benefits Statement
The World Bank is committed to preserving Saved:
endangered forests and natural resources. • 6 trees
The Office of the Publisher has chosen to • 1.7 million BTUs
print Skills, Not Just Diplomas: Managing of total energy
Education for Results in Eastern Europe • 534 pounds of net
and Central Asia on recycled paper with greenhouse gases
50 percent postconsumer fiber in accordance • 2,567 gallons of
with the recommended standards for paper waste water
usage set by the Green Press Initiative, a • 175 pounds of
nonprofit program supporting publishers in solid waste
using fiber that is not sourced from endan-
gered forests. For more information, visit
www.greenpressinitiative.org.
Restoring and sustaining growth in Eastern Europe and Central Asia requires reforms to
boost competitiveness and increase labor productivity. Among the required changes are
reforms to education. In surveys conducted immediately before the economic crisis,
companies in the region reported shortage of skills as one of the most significant bottle-
necks in their operations, suggesting that education systems in Eastern Europe and Central
Asia—with a reputation for high enrollment rates and well-trained teachers—still need to
improve their performance.
In fact, international test results show that many students—outside of a handful of coun-
tries in the region—are failing to acquire more than the most basic literacy and numeracy
skills. Anecdotal evidence also indicates that the rapid expansion in higher education has
led to a decline in the quality and relevance of education provided. At the same time, there
are few opportunities for adults to retrain, upgrade, or acquire new skills—the life-long
learning needed for employability.
As Skills, Not Just Diplomas suggests, the shortage of skills is a wake-up call to reform
education and training systems to provide higher quality education with the flexibility
for students and training institutions to better respond to market signals. Such deep
reform will have to center on the following:
• Focusing more on measuring whether students learn and graduates find jobs, and
using this information to actively improve teaching and learning.
• Using incentives across the education system, including granting greater autonomy
to institutions on curriculum, teaching methods, resource use and institutional mission,
and increasing accountability for learning.
• Addressing the huge inefficiencies in the sector by downsizing the school network
in response to dwindling student numbers, and using freed-up resources to restore
the attractiveness of the teaching profession, expand adult education and training, and
measure learning outcomes.
Only by embracing and seriously pursuing such difficult and long-term reforms will the
countries in Eastern Europe and Central Asia substantially improve their education outcomes
and support the competitiveness of their economic systems.
ISBN 978-0-8213-8096-3
SKU 18096

Skills, not just Diplomas

Автор:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Skills, not just Diplomas

Автор:

Авторское право:

Доступные форматы

DIREC TIONS IN DE VELOPMENT

Skills, Not Just Diplomas

Lars Sondergaard and Mamta Murthi

All rights reserved

Rights and Permissions

Library of Congress Cataloging-in-Publication data has been requested.

Chapter 1 The Demand for Skills in ECA 17

Wages Have Risen for Skilled Labor 31

Chapter 2 Education and the Supply of Skills to the

Chapter 3 Resolving the Skills Shortage in the

Chapter 4 Managing for Results at the Pre-University

Improve the Efficiency of Resource Use 126

Chapter 5 Managing for Results in the Tertiary

Chapter 6 Advancing Adult Learning in ECA 165

Chapter 7 Extended Summary: The Path for Education

7.3 Options for Making Schools More Accountable for

1.14 Employers’ Perceptions of Worker Skills as a

3.2 Learning to Use Data to Drive Education Policy 95

7.1 Assessing a Skill—Reading Performance in the

B5.3 Summary of Changes 150

It is no secret that good education lies at the heart of economic growth

sible, current spending on education needs to be used much more effec-

This study was managed by Lars Sondergaard, who authored chapters 3

AES Adult Education Survey

EBRD European Bank for Reconstruction and

ISCO-88 International Standard Classification of

UNESCO United Nations Educational, Scientific and

Country abbreviations used in figures and tables:

Note: These three-letter country codes are part of the International

When trying to answer these questions, this book faces a fundamental

The Skills Challenge

The ECA region is reasonably successful not only in terms of enroll-

Figure O.1 Analysis of Reading Competency of 15-Year-Old Students

4 BIH GEO EST

KOS KGZ HRV UKR

Source: Authors’ calculations based on BEEPS 2008.

be a major or severe constraint in 2008. The highest proportion of firms

Why Are Skills an Emerging Problem if Education

for their income level, ECA countries’ growth is constrained by skills

in international learning assessments at this level—which are likely the

agenda, particularly in more advanced ECA countries and countries that

Priority Areas for Action

Figure O.3 Primary School Student-Teacher Ratios in ECA Compared to

Managing Education Systems for Results

Focus Attention on Learning and Employment Outcomes

makers and higher education institutions because it helps them detect

Introduce Autonomy and Accountability Based on Results

improved learning outcomes to all types of students, including those

rankings, tracer studies, and creation of an enabling environment to

Improve the Efficiency of Resource Use through

Build the Foundations of Adult Learning Systems

The Demand for Skills in ECA

manual workers. As a result, returns to education and professional

Background: The Demand for Highly Skilled Labor

manual tasks, as well as nonroutine manual activities, has declined (Autor,

decentralized decision making, just-in-time operations, teamwork, and

Demand for Skilled Labor Has Risen in the ECA Region

Figure 1.2 Worker Responses to Survey on Working Conditions in Europe, 2005

Does your main job involve: (% of respondents answering "yes")

EU15 member states (i.e., non-ECA) new EU member states

Source: Authors’ calculations based on data obtained from Eurofound 2007.

occupation. The more educated a worker, the more skilled he or she is

of total job creation, exceeding the share common in developed countries

job creation job destruction